Hello!

I am recording midi tracks from motif XF into cubase, using an Sp/dif cable from the digital output into a firewire 400 input of my sound card.

The volume of the digital audio signal seems to be quite low, compared to the same tracks recorded with regular audio cables. The input volume is set to 0 db in cubase as well as on the fireface.
Is it normal??

Can I raise the input volume in the motif?

Thank you!!!

Saying the volume (sic) of the digital output is low, is basically saying your mix is low in level… Which is not necessarily a bad thing. A bad thing would be you exceeding 0dB.

Your mix output level is a sum total of your Parts.

You can read up on how and what affects your overall volume level in the following article:
MIDI Velocity and Audio Record Levels

Bad_Mister - 04 February 2014 06:43 AM

Saying the volume (sic) of the digital output is low, is basically saying your mix is low in level… Which is not necessarily a bad thing. A bad thing would be you exceeding 0dB.

Your mix output level is a sum total of your Parts.

You can read up on how and what affects your overall volume level in the following article:
MIDI Velocity and Audio Record Levels

Thank you.
I read the article also. What I still don´t understand is why the digital signal is so low COMPARED to the same tracks recorded with 2 audio cables.
I looked in “utility”, the audio out of the motif is set to + 6 db, but I couldn´t see any setting for digital out.
My mix levels are fine I think at least when I listen to my sequences through the motif.

Thank you

Digital audio does not output “volume” (not directly) - it is unaffected by your main Volume slider which sends analog audio (what we call “sound") to your analog output. The digital signal cannot be heard, its digital. It represents exactly your recorded levels.

It takes more signal than you are currently sending through the system to maximize the record level. This is a good thing because you can always increase the output level, as outlined in detail in the article.

Volume is subjective… A bigger, more powerful amplifier can make your mix playback at deafening levels. Volume is what we as humans perceive as loudness.

Gain or more commonly Level, is objective and is the signal as interpreted by a meter used to measure such things. The digital signal is measured by a meter. And represents your actual recorded signal. You cannot hear digital signal until it is converted back to analog.

Whether your recorded signal (level) tops out at -7dB or closer to 0dB on the meter will be perceived as two different degrees of loud, with the one closer to 0dB being louder (naturally)… But with the right analog playback system I can make them equally loud. Because you can turn up analog signal.

Even though one causes more movement of the meter, LOUD is subjective. I can still make them equally loud in the analog world… Because analog can be subjectively made louder.

To increase your recorded levels you will need to pay more attention to your individual contributors to your mix. Gain staging is one of the important things one must learn when recording audio.

One of the roles of the MASTERING ENGINEER is to take your mix that tops out at -7 and bring it up to ear-level where it will be perceived as being as loud as other mixes in the same genre.

Multi Band Compression is one of the “magic” tools found in the Mastering Engineers arsenal (it is also one of your overall Effects found in the XF’s Master Effect processor.

Try some of the Presets… Like the Maximizer, to see what type of affect this has on your overall mix. It goes so far as… You can take two mixes which both top out at -2dB on the meters and make one of them sound as much as 10 to 15dB louder, with the artful use of Multi-band compression. How the ear/brain interpret what us loud can differ greatly from what meters show you.

Musicians know this but don’t always apply the physics to what the intuitively know already.

A compressor can make us think something is louder without increasing the overall output level, by changing the relationship between soft and loud… Dynamic range.  Guitar players will tell you that a compressor will give you longer sustain.

Think about it - does that sound like its true? How does an audio device plugged inline with the electric guitar signal make the string ring longer?

Fact: it doesn’t.
But guitar players will swear to you that a compressor gives longer sustain. And when you hear it, you’ll most certainly agree. But if you get out your stop watch, the string stops vibrating at exactly the same length of time.

What the compressor does is flatten out the loud spike that peaks at the attack, so that the sound remains constant for more of the time it is sustaining - it stays at the same volume for a longer period than it would without compression. Funny as your brain starts to hear a sound fade down in volume, it will stop paying attention at a certain point. By flattening out the rise, initial fall, sustain, and release of the string, we perceive it as lasting longer.

As humans we say the compressed signal has more “presence”, more “weight”, it’s “fatter”, “thicker”, “richer”, more “in your face” - you’ve undoubtedly used these terms or something like them.

Even though the signal is now lower in overall level, you can, with a compressor, raise the overall output back to unity gain (its original level)… We have brought the softer sounds up closer to the loudest sounds, compressing the dynamic range… And although the level on the meter is the same (unity gain) you will swear that the compressed signal is significantly louder than the original.

Bad_Mister - 04 February 2014 08:37 AM

Digital audio does not output “volume” (not directly) - it is unaffected by your main Volume slider which sends analog audio (what we call “sound") to your analog output. The digital signal cannot be heard, its digital. It represents exactly your recorded levels.

It takes more signal than you are currently sending through the system to maximize the record level. This is a good thing because you can always increase the output level, as outlined in detail in the article.

Volume is subjective… A bigger, more powerful amplifier can make your mix playback at deafening levels. Volume is what we as humans perceive as loudness.

Gain or more commonly Level, is objective and is the signal as interpreted by a meter used to measure such things. The digital signal is measured by a meter. And represents your actual recorded signal. You cannot hear digital signal until it is converted back to analog.

Whether your recorded signal (level) tops out at -7dB or closer to 0dB on the meter will be perceived as two different degrees of loud, with the one closer to 0dB being louder (naturally)… But with the right analog playback system I can make them equally loud. Because you can turn up analog signal.

Even though one causes more movement of the meter, LOUD is subjective. I can still make them equally loud in the analog world… Because analog can be subjectively made louder.

To increase your recorded levels you will need to pay more attention to your individual contributors to your mix. Gain staging is one of the important things one must learn when recording audio.

One of the roles of the MASTERING ENGINEER is to take your mix that tops out at -7 and bring it up to ear-level where it will be perceived as being as loud as other mixes in the same genre.

Multi Band Compression is one of the “magic” tools found in the Mastering Engineers arsenal (it is also one of your overall Effects found in the XF’s Master Effect processor.

Try some of the Presets… Like the Maximizer, to see what type of affect this has on your overall mix. It goes so far as… You can take two mixes which both top out at -2dB on the meters and make one of them sound as much as 10 to 15dB louder, with the artful use of Multi-band compression. How the ear/brain interpret what us loud can differ greatly from what meters show you.

Musicians know this but don’t always apply the physics to what the intuitively know already.

A compressor can make us think something is louder without increasing the overall output level, by changing the relationship between soft and loud… Dynamic range.  Guitar players will tell you that a compressor will give you longer sustain.

Think about it - does that sound like its true? How does an audio device plugged inline with the electric guitar signal make the string ring longer?

Fact: it doesn’t.
But guitar players will swear to you that a compressor gives longer sustain. And when you hear it, you’ll most certainly agree. But if you get out your stop watch, the string stops vibrating at exactly the same length of time.

What the compressor does is flatten out the loud spike that peaks at the attack, so that the sound remains constant for more of the time it is sustaining - it stays at the same volume for a longer period than it would without compression. Funny as your brain starts to hear a sound fade down in volume, it will stop paying attention at a certain point. By flattening out the rise, initial fall, sustain, and release of the string, we perceive it as lasting longer.

As humans we say the compressed signal has more “presence”, more “weight”, it’s “fatter”, “thicker”, “richer”, more “in your face” - you’ve undoubtedly used these terms or something like them.

Even though the signal is now lower in overall level, you can, with a compressor, raise the overall output back to unity gain (its original level)… We have brought the softer sounds up closer to the loudest sounds, compressing the dynamic range… And although the level on the meter is the same (unity gain) you will swear that the compressed signal is significantly louder than the original.

Ok, now it´s far more clear to me! I didn´t want to use compression before mixing, but I will probably raise all the XF mixer sliders to 127 in order to gain level. Or just record from the analog outputs instead of digital.
Thank you for your time again!

Ok, now it´s far more clear to me! I didn´t want to use compression before mixing, but I will probably raise all the XF mixer sliders to 127 in order to gain level. Or just record from the analog outputs instead of digital.

Please learn to use the VCE EDIT function accessible from the [MIXING] screen. (I describe this in detail in the linked article)
Drop into Edit on each Voice… Raise its VOICE VOLUME to 127 if necessary.
Select a Part
Press [F6] VCE EDIT
Press [COMMON EDIT]
Press [F1] GENERAL
Press [SF2] OUTPUT
Adjust the Voice Volume
Press [STORE]
Store your newly edited Voice to the MV (Mixing Voice) location within this Song/Pattern
Rinse and repeat.

(drum kits will need to be stored in the User Drum bank)

Then when you return to the Mixer use your faders to lower the levels of each PART to create a musical balance of the instruments.

The difference between the VOICE VOLUME (source) and PART VOLUME (at the mixer) needs to be understood.

If you’re the recording engineer and an incoming signal does not give you the LEVEL you require, even when your fader is all the way up, you would ask the musician to turn up the signal at the source.

Make sense? This allows you to use the fader to lower the levels so you get a balanced mix that is gain staged properly for your MIXDOWN!

Bad_Mister - 04 February 2014 10:22 AM

Ok, now it´s far more clear to me! I didn´t want to use compression before mixing, but I will probably raise all the XF mixer sliders to 127 in order to gain level. Or just record from the analog outputs instead of digital.

Please don’t raise all of the mixer sliders to 127… That is the antithesis of mixing. The sliders, called Faders, are called that so that you can turn down (fade) the output level of each Part to make a musical balance, a mix.

Please learn to use the VCE EDIT function accessible from the [MIXING] screen. (I describe this in detail in the linked article)
Drop into Edit on each Voice… Raise its VOICE VOLUME to 127 if necessary.
Select a Part
Press [F6] VCE EDIT
Press [COMMON EDIT]
Press [F1] GENERAL
Press [SF2] OUTPUT
Adjust the Voice Volume
Press [STORE]
Store your newly edited Voice to the MV (Mixing Voice) location within this Song/Pattern
Rinse and repeat.

(drum kits will need to be stored in the User Drum bank)

Then when you return to the Mixer use your faders to lower the levels of each PART to create a musical balance of the instruments.

The difference between the VOICE VOLUME (source) and PART VOLUME (at the mixer) needs to be understood.

If you’re the recording engineer and an incoming signal does not give you the LEVEL you require, even when your fader is all the way up, you would ask the musician to turn up the signal at the source.

Make sense? This allows you to use the fader to lower the levels so you get a balanced mix that is gain staged properly for your MIXDOWN!

Yes it makes sense!

benoit - 04 February 2014 10:03 AM

Ok, now it´s far more clear to me! I didn´t want to use compression before mixing, but I will probably raise all the XF mixer sliders to 127 in order to gain level.

Rather than raising all the levels to 127, there’s an approach that will keep the mix balance about the same, by preserving the ratios of the levels close to what they were originally.

For example, if the highest level is 110, and you bring it up to 127, it will be raised by 127/110, or about 1.16 times. If you then raise all the remaining levels by that 1.16 multiplier, the levels will obviously all be higher, but the ratios will be nearly what they were previously. It’s not exact - the relationship between the 0-127 scale and the levels isn’t linear - but it’s usually close enough unless one of the levels is very low to begin with.

Of course, you might need a calculator…  :-)

5pinDIN - 04 February 2014 06:46 PM

Rather than raising all the levels to 127, there’s an approach that will keep the mix balance about the same, by preserving the ratios of the levels close to what they were originally.

For example, if the highest level is 110, and you bring it up to 127, it will be raised by 127/110, or about 1.16 times. If you then raise all the remaining levels by that 1.16 multiplier, the levels will obviously all be higher, but the ratios will be nearly what they were previously. It’s not exact - the relationship between the 0-127 scale and the levels isn’t linear - but it’s usually close enough unless one of the levels is very low to begin with.

Of course, you might need a calculator…  :-)

I thought the sliders were representing logarithmic scale as is the case with dB? Or are they linear so as to allow the use of simple ratios as suggested by you?

Forgive my ignorance! :-)

EDIT: On a 2nd thought, it seems that the use of simple ratios to maintain the overall balance in the mix might just be the appropriate thing to do because the sliders use a logarithmic scale as do our ears for the perception of loudness! I guess I’m completely confused now. ;-)

Hi VikasSharma - we missed you, and I’m glad that one of my posts has encouraged you to post. I tried to avoid the math, but I’ll explain…

Yes, our hearing is approximately logarithmic, as is the level change over the 0-127 range. (I did say ”the relationship between the 0-127 scale and the levels isn’t linear”.  :-) )

I also said ”but it’s usually close enough unless one of the levels is very low to begin with”. That part of my statement recognizes two things:
1) The log function is quite non-linear at small values.
2) The log function “straightens out” considerably at higher values.
Plots of the log function are easily found online, so I won’t attach one or provide links.

From a practical standpoint, I’ve found that simple ratios work very well on levels in the 60-127 range, and quite decently down to about 30. Below that, of course, the log curve has an increasing effect, so the ratio method might require some “tweaking” of lower value levels in the mix. It’s been my experience, though, that most mixes don’t have too many levels set low enough for that to be much of a concern.

Under any circumstance, I think my approach is better than unbalancing the mix by ramming everything up to 127. Of course, it doesn’t work if one of the levels is already at or very near 127.  ;-)

I can cancel my Berkeley College of Music course now… you all are really good!

The National Recording Arts and Sciences as well as the best Philosophy professors would enjoy your analyzations!  :)

*I do not OWN the photo*

THIS is why I give a little ‘coin’ to Wiki :)

Click thumbnail to see full-size image

No math is needed to mix… It is an aural art. Math never hurts, but none is necessary here.

The Voice Volume is GAIN
The Part Volume is where you mix.

The Voice Volume setting is the input source
The mixer is where you access Part volumes (using the CS) these adjust PART volume, and is where you subjectively mix.

Need more signal from a channel, increase the gain. For subjective balancing of instruments (the “musical” mix) adjust the Part Volume.

Leave the math to the mathematicians (in this case).

I think Benoit, the OP, got the message.

Bad_Mister - 05 February 2014 05:57 AM

No math is needed to mix… It is an aural art. Math never hurts, but none is necessary here. [...]

I didn’t say math was “necessary”. What I offered is an approach that works when you already have a good mix balance.

 

Bad_Mister -

Leave the math to the mathematicians (in this case).

Obviously, the final product should satisfy ears, not strictly be dictated by a formula.

 

Bad_Mister -

I think Benoit, the OP, got the message.

I’m confident that most forum members are capable of deciding for themselves how to deal with the situation.

5pinDIN - 05 February 2014 07:40 AM

…an approach that works when you already have a good mix balance.

Exactly. :-)
Thank you!

Bad_Mister - 05 February 2014 05:57 AM

No math is needed to mix… It is an aural art. Math never hurts, but none is necessary here.

The Voice Volume is GAIN
The Part Volume is where you mix.

The Voice Volume setting is the input source
The mixer is where you access Part volumes (using the CS) these adjust PART volume, and is where you subjectively mix.

Need more signal from a channel, increase the gain. For subjective balancing of instruments (the “musical” mix) adjust the Part Volume.

Leave the math to the mathematicians (in this case).

I think Benoit, the OP, got the message.

I find this so good that I will write it on the motif manual.
Some voices, however, are already originally set to 127 volume.
I think that a combination of voice volume setting/ and/ or part volume ajustement in mixing mode should be sufficient to have enough gain.

5pinDIN - 04 February 2014 06:46 PM

benoit - 04 February 2014 10:03 AM

Ok, now it´s far more clear to me! I didn´t want to use compression before mixing, but I will probably raise all the XF mixer sliders to 127 in order to gain level.

Rather than raising all the levels to 127, there’s an approach that will keep the mix balance about the same, by preserving the ratios of the levels close to what they were originally.

For example, if the highest level is 110, and you bring it up to 127, it will be raised by 127/110, or about 1.16 times. If you then raise all the remaining levels by that 1.16 multiplier, the levels will obviously all be higher, but the ratios will be nearly what they were previously. It’s not exact - the relationship between the 0-127 scale and the levels isn’t linear - but it’s usually close enough unless one of the levels is very low to begin with.

Of course, you might need a calculator…  :-)

Thank you I appreciate your comments as well.
I thought about doing this also, but I thought that there should be enough gain by keeping the sliders around 100 as a starting point for an internal mix.
And yes, I need a calculator…