Cubase SX 2 cuts high frequencies?

[mauseoleum]

Topic: Cubase SX 2 cuts high frequencies?

Karl has a deal with mastering eingineers so they don't loose their jobs (and buy Wavelab). So ... Cub comes with hi-freq scisors built-in.

[mauseoleum]

see, everything happens for a reason (i should sell this slogan to props instead whining on forums).

[stefancrs]

Thanks all for your contribution to my problem!
I don't think there is a problem with my setup.
I have Echo MIA soundcard connected with balanced cables to TR6 event monitors.
I set the project in cubase to 44/16. So i suppose
that anything i hear when i'm playing back the mix(before exporting) is 44/16 (vsti's and audio files).
But if cubase is using 32bit internal then i should choose to export in 32bit WITHOUT dithering for further editing in soundforge (mastering) and then the last step in soundforge would be a dither plugin and export to 44/16 in order to burn the track to the cd. If that's the case then maybe if i export individual tracks (to save cpu) i should export them in 44/32bit?

What you hear is probably not due to a problem at all.

This again?
SX2 does not cut high frequencies during mixdown or anything.
You probably play the mixdown externally at a different volume than what your arrangement played in.

I don't think you have to work in 32bit, because the only difference would be that you record your audio in 32bit. 16 is enough if you're careful when recording, 24 makes recording a bit easier.

[Manc Chris]

Confused.

Ok, I use SX myself. Normally record midi (vsti)and audio in a 48/24 project. Mixdown the master to 48/24 no dither, than master in soundforge and dither to 44/16.

After all this debate, can someone advise a better project setup?

[MickGael]

When I export a track from cubase the mix sounds duller somehow lacking high frequencies. I tried equallizing but it turns out that i over-equallize things and even if i do that the spectrum analyzer in soundforge (or the plain simple freq analyzer on winamp) doesn't show "enough" high frequencies!

Hey Kouts,
Nope - it's nope just you.

There have been several posts at the Steinberg forum (as well as many people claiming there is no change in the sound).

I am a trained engineer, and there absolutely is a change in the sound if you record at 24/44.1 - even if you mix to 24 bits.

The high-end becomes a tad muted, the stereo field collapses a bit, and in general the sound becomes pinched, strained, narrow and veiled. Granted, these are a matter of degree, but as far as I am concerned it is documented.

I have never been able to make a mix sound as good as it does when I am simply playing a mix in SX2. The only workable solution I have found is to mix to 32-bit. Even then, I invariably have to add something to the top end and work to "restore" some of the stereo field.

Quite maddening, especially when some people try and tell me that I can not possibly be hearing what I amost definitely do hear. Even more, Steinberg has no comment.

[IIRs]

Confused.

Ok, I use SX myself. Normally record midi (vsti)and audio in a 48/24 project. Mixdown the master to 48/24 no dither, than master in soundforge and dither to 44/16.

After all this debate, can someone advise a better project setup?

Either stick to 44.1KHz throughout, or go to 88.2KHz: any advantages of 48 over 44.1 will be negated by the SRC at the end.

Export a 32-bit file for mastering, or if you stick to 24-bit, add dither.

[Spaceman Sounds]

This is true, myself and many others have found this. The solution is to use a plugin like the Voxengo free recorder, that you use as an insert on the master fx and record the mix as you play.

I have found a big difference in what is exported and what you hear when playing Sx 2 or Nuendo 2 back live.

[stefancrs]

There have been several posts at the Steinberg forum (as well as many people claiming there is no change in the sound).

I am a trained engineer, and there absolutely is a change in the sound if you record at 24/44.1 - even if you mix to 24 bits.

The high-end becomes a tad muted, the stereo field collapses a bit, and in general the sound becomes pinched, strained, narrow and veiled. Granted, these are a matter of degree, but as far as I am concerned it is documented.

The bit depth you record in will affect the sound quality if you in any way want to alter the recorded audio, which you most likely will. So this is correct.

I have never been able to make a mix sound as good as it does when I am simply playing a mix in SX2. The only workable solution I have found is to mix to 32-bit. Even then, I invariably have to add something to the top end and work to "restore" some of the stereo field.

Yes, you need to add volume, because the mix plays a bit higher within cubase than in mixdown.

Quite maddening, especially when some people try and tell me that I can not possibly be hearing what I amost definitely do hear. Even more, Steinberg has no comment.

Sure it is audible, but it is proabably just a difference in volume. Loss in volume will be perceived as dullness and lack of clarity, worse stereo separation and so on. It's not the actual audio, it's just your ears.

[stefancrs]

This is true, myself and many others have found this. The solution is to use a plugin like the Voxengo free recorder, that you use as an insert on the master fx and record the mix as you play.

I have found a big difference in what is exported and what you hear when playing Sx 2 or Nuendo 2 back live.

Volume.

[SJ_Digriz]

You guys keep saying you need to use "dither" to keep "quality". Dithering creates NOISE to limit the amount of percieved truncation artifacts (digital harshness). How and where the noise is placed is based on the particular tool. It really has nothing to do with quality sound. In fact you will find a boat load of engineers in acoustic studios who would shoot you for using a dithering tool instead of recording/mixing properly.

[snooky]

You guys keep saying you need to use "dither" to keep "quality". Dithering creates NOISE to limit the amount of percieved truncation artifacts (digital harshness). How and where the noise is placed is based on the particular tool. It really has nothing to do with quality sound. In fact you will find a boat load of engineers in acoustic studios who would shoot you for using a dithering tool instead of recording/mixing properly.

????

ok, that's just ignorant...any engineer worth it's salt would have to use dithering to keep highres projects sounding good when downsampling to lowres CD audio.

[C00kie]

Confused. Normally record midi (vsti)and audio in a 48/24 project. Mixdown the master to 48/24 no dither, than master in soundforge and dither to 44/16.

After all this debate, can someone advise a better project setup?

Yes, only one minor improvement: record & run the project in 44/24, so you don't need to resample from 48 to 44.1kHz but only dither at the end.

Probably not much quality difference, but you'll gain the time waiting for resampling plus about 10% more CPU power available in Cubase.

[IIRs]

You guys keep saying you need to use "dither" to keep "quality". Dithering creates NOISE to limit the amount of percieved truncation artifacts (digital harshness). How and where the noise is placed is based on the particular tool. It really has nothing to do with quality sound. In fact you will find a boat load of engineers in acoustic studios who would shoot you for using a dithering tool instead of recording/mixing properly.

Bullshit! Dither will increase the apparent resolution of the reduced wordlength signal, while reducing quantisation distortion across the whole dynamic range. I suggest you do some reading on the subject.

[SJ_Digriz]

There is no loss of highs. If you export bit depth to the same bit depth you will have a slight loss of db. This is true in any application.

When you CHANGE bit depth there are many factors that can lead to this loss of highs. The first thing you can do is do your summed exports at the same bit depth. You will have only marginal db loss this way.

If you are already at 16bit and exporting summed to 16 you will have a file that is probably close to 4->6 db quieter than the monitored level. To overcome this people start jacking with master strip compressors and eqs. THIS is what leads to the loss of sound quality.

Soundforge/Wavelab usually don't exhibit this behavior because the track you are exporting is ALREADY summed. You are working with a single audio stream. I've done the tests where you export a given 44/24 track from Wavelab and Cubase and do nothing but go to 44/16. Their is no difference in the output. They both put out an identical file.

[SJ_Digriz]

Bullshit! Dither will increase the apparent resolution of the reduced wordlength signal, while reducing quantisation distortion across the whole dynamic range. I suggest you do some reading on the subject.

I love the read reference..good one.

You need to read closer however hehehe. You can't increase the resolution. You nicely put apparent resolution. Dither works many different ways so I won't make a blanket statement. But increase resolution is NOT one of them. It replaces one kind of noise with another. It also, sometimes as an option, sometimes without option, puts noise in the silent spots.

Again, there is very minimal truncation distortions if the tracks are recorded, gain staged, summed and exported with skill. The dither actually does more harm than the original truncation errors.