How to calibrate levels on the Focusrite ISA 2

[armans]

Can someone post a link to the Claret 4 TBs specifications please?
I'll be jiggered if I can find it!

Dave.

It was replaced by the new usb one. I have the manual at home but even there it does not state the max input for the back inputs. Focusrite only put the max input specs for the front inputs into the manual. I wonder why, maybe because they prefer not to show it.

[armans]

So, unless the correct figures can be found we could be doing Focusrite a grave injustice?

+10dBu seems a very poor headroom figure for even a sub £100 interface.

From focusrite:

If you have the Clarett 4Pre USB then you are correct, the maximum input is +18dBu, but if you have the Clarett 4Pre this is +10dBu.

[armans]

Yes, got that Hugh but it states "Max input +26dBu" Where is the specc for the fixed inputs at +10dBu?

Maybe they left it out on purpose?

[armans]

I've no idea where the idea of a +10dBu max line input came from. It seems very odd to me.

All the hardware facilities and other input and output specs appear identical between the USB and TB models, so I'd have expected the rear panel line inputs to have the same, entirely reasonable, +18dBu max input level.

H

I purchased 2 X -15db in-line attenuators. I received them today and was able to calibrate the meters using the back inputs but I must say I am rather dissapointed at how useful the meters will be anyway. What you said about different meters showing different things must be the reason. When I use a steady tone the meters show one thing but when I actually play my guitar and record, the meters in my DAW show much higher peak levels than what the ISA 2 shows me. I think my DAW shows peak values while the Focusrite shows average levels because the meter is always much lower than what I am recording

[Hugh Robjohns]

When I use a steady tone the meters show one thing but when I actually play my guitar and record, the meters in my DAW show much higher peak levels than what the ISA 2 shows me.

Welcome to the world of 'standards'... We all love standards in the audio industry. That's why we have so many of them... all different...

A meter is just an indicator, and with experience you get to learn how to interpret what different meters are indicating.

So you just need to learn what an acceptable peak level is on the ISA meter and reference to that.

...and remember that working with a generous headroom margin always makes life much easier when recording as you don't need to worry about peak levels when you know nothing you do will cause an overload. In that situation, the meter is really just confirming the signal is present and correct.

[armans]

When I use a steady tone the meters show one thing but when I actually play my guitar and record, the meters in my DAW show much higher peak levels than what the ISA 2 shows me.

Welcome to the world of 'standards'... We all love standards in the audio industry. That's why we have so many of them... all different...

A meter is just an indicator, and with experience you get to learn how to interpret what different meters are indicating.

So you just need to learn what an acceptable peak level is on the ISA meter and reference to that.

...and remember that working with a generous headroom margin always makes life much easier when recording as you don't need to worry about peak levels when you know nothing you do will cause an overload. In that situation, the meter is really just confirming the signal is present and correct.

Well said thanks for all your help. I just have one last question regarding this. Lets say, I never used in-line attenuators and just used less gain on my ISA 2 so I could go into the back if my Clarett interface. I would obviously use 14db or so less gain going into my interface. Would there be any problem doing this? I mean I was doing it all along until I realised something was not right with the meters but does it afffect the sound in some way?

[Hugh Robjohns]

Lets say, I never used in-line attenuators and just used less gain on my ISA 2 so I could go into the back if my Clarett interface. I would obviously use 14db or so less gain going into my interface. Would there be any problem doing this? I mean I was doing it all along until I realised something was not right with the meters but does it afffect the sound in some way?

It would mean running the ISA2 preamp with less gain and a lower average level and therefore, in theory at least, with a reduced signal-to-noise performance.

In practice, though, I doubt you'd actually notice any real difference because we're only talking about 8dB or so.

H

[armans]

In practice, though, I doubt you'd actually notice any real difference because we're only talking about 8dB or so.

H

ISA max output 24db
Clarett 4 max input 10db

Isnt that a 14db difference?

[Hugh Robjohns]

You and your numbers...

I was thinking of the +18dBu max input capability of the USB version of your interface, as opposed to the +10dBu max of the TB version you have. That's 8dB by my reckoning...

And if you ever did get +24dBu out of your ISA you've really been caught out badly ... HEADROOM, remember?

[armans]

You and your numbers...

I was thinking of the +18dBu max input capability of the USB version of your interface, as opposed to the +10dBu max of the TB version you have. That's 8dB by my reckoning...

And if you ever did get +24dBu out of your ISA you've really been caught out badly ... HEADROOM, remember?

Sorry to come back to you again but there is still a problem with my ISA meters.

I tried a white noise test and then the problem became evident. This is precisely what I did.

I sent a -14dbfs white noise test tone out of output1 of the CLarett and into the ISA 2 line input.
I set the ISA 2 gain on the ISA 2 to 0 and sent it back into the Clarett using input 5 going through a -15db pad (the pad is advertised as being 15db but not sure it it)
I looked in my DAW and the tone coming in was 4db hotter than the output in my DAW and showed a level of -10dbfs
I then tweaked the meter on the back of the ISA 2 and even in the most anti-clockwise position only the -18dbfs light lit up. I was at least 8db short to make the calibration happen.

So can you please help me make sense of these results?

[Hugh Robjohns]

Sorry to come back to you again but there is still a problem with my ISA meters.

No... in this case the problem is with your lack of understanding of metering as a whole...

I tried a white noise test and then the problem became evident.

You CANNOT use any type of noise to 'calibrate' an audio meter, or to check levels. Noise is a random signal which is constantly changing in level over time, and specifically the peak level is constantly changing with respect to its RMS level. So different types of meter will always provide completely different readings.

You can only check and calibrate audio meters using steady-state sine tones -- 1kHz is normal, but 400Hz is easier on the ear!

[armans]

No... in this case the problem is with your lack of understanding of metering as a whole...

That too ! but how much could the noise signal "fluctuate"? I see it isn't steady in my DAW but it hardly varies more than a db or so. So how can the results be so different? Even when I go straight out of my Clarett output 1 straight into output 5 with the noise, there is a 5-6 db difference in the level compared to when I am using a sine tone.

You can only check and calibrate audio meters using steady-state sine tones-- 1kHz is normal, but 400Hz is easier on the ear!

Ok good to know, thanks!

[ef37a]

No... in this case the problem is with your lack of understanding of metering as a wholes...

That too ! but how much could the noise signal "fluctuate"? I see it isn't steady in my DAW but it hardly varies more than a db or so. So how can the results be so different? Even when I go straight out of my Clarett output 1 straight into output 5 with the noise, there is a 5-6 db difference in the level compared to when I am using a sine tone.

You can only check and calibrate audio meters using steady-state sine tones-- 1kHz is normal, but 400Hz is easier on the ear!

Ok good to know, thanks!

Mind Hugh! For the craic I once connected a £1000 ac mVoltmeter, a Fluke true rms meter and a 30 quid Maplin DMM to a white noise generator. They all gave different readings by around 3dB iirc!

Dave.

[Hugh Robjohns]

...how much could the noise signal "fluctuate"? I see it isn't steady in my DAW but it hardly varies more than a db or so. So how can the results be so different?

If you really want to know, buy and read this book:

https://www.soundonsound.com/reviews/audio-metering-measurements-standards-practice-second-edition

Even when I go straight out of my Clarett output 1 straight into output 5 with the noise, there is a 5-6 db difference in the level compared to when I am using a sine tone.

Yep. Sounds about right. It's partly to do with the Crest Factors of different signals, and partly the different meters' integration times and general ballistics -- the latter because unlike a steady tone, a noise signal is inherently changing its amplitude continuously.

The Crest Factor is the difference between a signal's true peak amplitude and its RMS value. The Crest factor for a sine wave is fixed an steady at 3dB, but it can be very much higher and variable for a noise signal... or a music signal for that matter.

So if one meter is trying to indicate something akin to peak values, and the other is trying to show something like the RMS values, you will inevitably have a substantial difference in readings.

And similarly, if one meter has a very fast response (short integration time) -- like a peak-reading meter -- but the other has a slower response -- typical of an RMS meter -- you will also get very different readings as one will react well to instantaneous level changes in the noise (or music) signal, and the other won't even see them and will only react to longer term amplitude changes ....

H

[armans]

And similarly, if one meter has a very fast response (short integration time) -- like a peak-reading meter -- but the other has a slower response -- typical of an RMS meter -- you will also get very different readings as one will react well to instantaneous level changes in the noise (or music) signal, and the other won't even see them and will only react to longer term amplitude changes ....

H

Ok thank you for clarifying but I still have to say that there is something not quite right about this. As another test, I played a pad on a VST instrument and once again, the pad showed a hotter return level in my DAW than what the ISA showed on its meter so as it stands, the meter is of no use to me at all! It is about 5 - 8db off and at focusrite they assured me that peak RMS values have nothing to do with it. They confirmed that what the meter shows on the ISA is what should be in my DAW. They are busy helping me but they only reply once a week so it is a tedious process. I will of course let you all know how it is going but for now I stand firm that the ISA meter does not work correctly. If there is such a difference in level between the sine tone and white noise / pad instrument then it would be more accurate to calibrate to those levels than to the sine tone because that way when I play and record something the level will be more like those levels and not like the sine tone, do you follow?

[Hugh Robjohns]

Ok thank you for clarifying but I still have to say that there is something not quite right about this.

I don't know how to make it any clearer.

No two different meters on different devices will ever show exactly the same readings on programme audio (anything other than steady test tones). They might get pretty close if they are the same types of meters, but even then the chances are some different material will still reveal differences.

If you want identical readings you have to have identical meters. That's why every studio across the BBC uses exactly the same type of PPM built to exactly the same specifications.

In your case, you have a virtual software meter and a hardware meter, and we don't actually know the specifications for either of them. However, your tests have already revealed that they are different types of meter with very different characteristics.

...as it stands, the meter is of no use to me at all!

As I said, all you have to do is familiarise yourself with the kind of acceptable readings displayed on each type of meter for normal recordings.

I used to work with tape recorders having VU meters on them, and mixing consoles with PPMs. Very different displays with different material, but I learned to interpret them perfectly well and make reliably consistent recordings using either. I'm sure you can learn to do the same -- and you have the added advantage of being able to tweak the calibration of the ISA meter to make it display something better suited to your own specific requirements...

Focusrite they assured me that peak RMS values have nothing to do with it. They confirmed that what the meter shows on the ISA is what should be in my DAW.

Very clearly, that statement is a nonsense and either you or they didn't fully understand the situation being discussed.

...for now I stand firm that the ISA meter does not work correctly.

You could, of course, be right... but I very much doubt it. I believe it's simply that you have two different meters but expect them to behave identically to music programme... which simply can't and won't ever happen.

If there is such a difference in level between the sine tone and white noise / pad instrument then it would be more accurate to calibrate to those levels than to the sine tone...

You can't because, as I have explained several times, different material has different Crest Factors and so different material -- different instruments, different keyboard presets -- will always give different -- and unpredictable results.

Instead of getting bogged down and fixating on trying to obtain identical readings, can I suggest you simply figure out what would an acceptable recording level as displayed on each kind of meter. When tracking, the absolutely peak levels really aren't critical as long as you're working with a decent headroom margin. They could be anywhere between -20dBFS and -6dBFS without any problem whatsoever. You can then tweak the recorded tracks to suit the mix as necessary without any issues at all.

Moreover, if you keep your monitoring levels consistent, you'll soon learn to set recording levels entirely by ear without ever looking at a meter, and probably do so with an accuracy of decibel or two quite easily.

[ef37a]

Armans, I can give you two examples of the problems I faced with 'meters'.

Many years ago I made peak reading LED device to fit inside my quite decent Yamaha cassette machine. I was not happy with the mechanical VU meters. I set the LED to just start to glow at 3dB over Dolby level*.

Over the next few weeks I was recording a variety of things, mainly R3 and some vinyl for the car. I noticed that very often, although music was not that loud, the LED blinked even though the VU meters were hardly moving (BBC bogs called then "Virtually Useless"!) I checked the calibration several times over a month or so but it was always spot on. The fact was a short crack on a snare would light the LED but not move the needles. That is about as extreme an example of two different meters not agreeing as you can get but as Hugh has pointed out several times (sigh) only identical meters will track on programme or noise.

A similar situation arose only a few years ago when my son was trying to even out the loudness of some 4 hours of music, mainly punk, for his sister's 40th birthday party.
He tried to do this in Samplitude, Sonar, Cubase and Audacity but no meter system was consistent from song to song. In the end he had to rely on the Mk 1 lugs!

Just thank your stars you are 'digital' and have that MOOOhassive dynamic range to play with!

*YOU! Will have to Google that if you want to know but it is water gone well under the pont now.

Dave.

[armans]

thanks for all the help and for sharing your experiences. I will report back if I have any breakthrough

[Hugh Robjohns]

[armans]

You CANNOT use any type of noise to 'calibrate' an audio meter, or to check levels.

From the Sweetwater website.

https://www.sweetwater.com/insync/pink- ... medium=CPC

Pink noise is the right type of noise to use to calibrate audio equipment (at least if you are using it for equalization calibration). Here’s where everyone gets confused:

Is calibrating meters so different?

[Hugh Robjohns]

Is calibrating meters so different?

Yes... for all the reasons I've already explained.

Pink noise (not white noise) can and usually is used for setting levels from monitoring speakers -- which is what that article is fundamentally referring to (room/speaker equalisation). It's also a handy reference sound when setting initial mix levels, and it's even quite good for masking background noises when you're trying to get to sleep...

But pink noise cannot be used as a reference level for accurate meter alignment.

And actually, that article is rather confused and inaccurate anyway, BTW...