Sound quality

A null test is totaly different from comparing files that are created in different versions of Cubase using it’s native processing.

There are two things going on here; you have the " null test people" that will keep on telling you that you are wrong and imagining things, and the people that just hear a difference when they open a project created in another version of Cubase.

Shure, null tests WITHOUT processing obviously always should null (duh!). But that is not what these posts are about I think…

People expect that after an update, everything sounds the same when they open an old project. Suppose the eq has been altered, but you do not use the eq. You would not notice the difference. But you would if you used it on a few tracks, right?

I had this experience with files in which I had used the cubase eq just on snare and bass, but I noticed something was drasticly different, not as intended right away (using cubase 7 vs 6.5). I just know my mixes. It took me some time to find out the difference was in the eq, using the same procedure as with a ‘null test’.

My experience sofar has been that due to various and constant changes in the software in different versions of Cubase (native eq, filters, group summing), the processed mixes do sound different in different versions, and they do not null at all If you happen to be using Cubase features that have changed in the update.

This means that Cubase is not fully compatible when using different versions, so you better not switch versions while working on a project in witch you use any of the native features of cubase, or you’d just have to accept to have to rebuild your mix a little…

I decided to add cubase version info to my files, and I have 3, 6, 7 and 8 installed for compatibility…

Rebuilding mixes in newer versions sometimes gave better results, sometimes I just gave up…

You could consider to only use 3rd party plugins and avoid using Cubase processing.

It would be convenient if Steinberg would provide a list of incompatibilities between different versions…

I personaly do not mind if they implement better eq, filters or group summing, but I would not mind to be notified about it in a better way…

Just discovered another gem.
In the Functions Menu of the Mixer (Arrow at top far right) there is a setting “EQ/Filter Transition” set to Soft by default but can be set to Quick. Mentioned on Pg. 295 of the Operation Manual, but nowhere an explanation of what it actually does. This function I believe was introduced with Cubase 7.

I noticed this in my mix as well. I also think C8 is is processing C7.5’s EQ differently. I opened a track i made with 7.5 into 8, and it sound very different. Much more bass, even if all the settings are exacly similar. Also, for some reason one of the instrument tracks is playing much louder in C8 than in C7.5. I have my bassline set -3db in C7.5, in C8 i have to set that same bassline to -12db for it to be in same volume levels. Weird.

Hmmm…

Any solution to Pre EQ thingy yet? I noticed that other people are also having problems when they open C7.5 project in C8 and the Pre EQ filters are messed up?

I’m still stuck with all my projects in C7.5 because what comes back in C8 has NOTHING to do with what the mix was before.

This is the worst f-up I’ve ever seen from Steinberg in the 20 years I’m using their products!

I’ve sent Steinberg support a test project to acknowledge the issues, but they decided to ignore it.
The Filters are off with the slope and frequency. Even with the slop set to the closes possible, the I have to approx half the frequency to get in the ballpark. So a HPF set to 11kHz need to be changed to 5.5kHz to line up. But no matter what I do, changing all slopes and frequ. combinations, I can’t get them to sound the same. I can’t get them to null with phase at 180degrees.
I also looked at the LPF and here we have the same problem. A filter set at 100 has to be moved up to 400-500Hz(!!) to get in the ballpark.

If ProTools would do such a thing they would be killed from the studios! I don’t know how Steinberg can get away with this.

The problem is still present in 8.0.20.

You can very easily make a test:

Begin in C7.5.
Use a 4 bar loop of white noise (test signal generator) and add a HPF/LPF. Then print the 4 bar loop with the LPF/HPF as audio track. When you reverse the printed track 180 degrees, both tracks null out perfectly.

Now open the project in C8. Right away, when you hit play, the tracks don’t null anymore at all. Now check by how far you have to adjust the HPF/LPF to get in the ballpark of the printed audio tracks. Your jaws will drop!

I contacted them. Sent them a test project with black and white evidence that the filters where off by miles. Mixes coming back totally different. No response. Telephone support denied everything. That was 6 months ago.

The evidence:
Here is the test project with white noise. Open it in C7.5 an hit play. Perfect Null.
open it in C8 and everything is off.
I recorded three different frequencies. 2x HPF and 1x LPF.

Download link:
https://dl.dropboxusercontent.com/u/32152722/Cubase%208%20filter%20bug.zip

I have the feeling, that even if i disable the elastiqe stretch in the audiopool, the files i imported from other people still get processed somehow since CP8. It’s like abelton. I really hope that this is a bug. And i’m still waiting for the new elastiqe v.3.x to be implemented in CP8. It’s out for over a year. I got the feeling, that the 2.x algorithm is so deeply integrated in CP8, that the programmers really have trouble to switch…

I use the Elastique algorithm so little that I can’t comment from a wealth of experience. But I do use it on occasion.

My use tends to be in very minute chunks, i.e. maybe a piece that has some very minor timing issues (off by a few milliseconds). I’ll use the cut tool to delineate the start and end, then time stretch to adjust the timing. Then I’ll switch back to “regular drag” and move some of the outside parts in order to overlap a very minor amount and use crossfade to blend the parts together.

I’ve never had any issues with it in this use case.

Since larger timing adjustments are usually indicative of timing problems throughout the track, anything more than minute stuff usually requires that I load up the track in Melodyne.

Yeah, the problem is still there. I am currently still using 7.5 because all the mixes sound very wrong when i open them at Cubase 8. It’s the filters. No matter how i set the slopes, etc. cant just make it sound same (or just too lazy to spend hours getting them right since they are perfectly right in C7.5).

I use LPF and HPF almost in every instrument track and now all those tracks that use LPF / HPF (or both) are messed up.
Updated to latest version, still no help.

I really want to move using C8 because i paid for it. But now i can’t because the filters are messed up when i open project from 7.5.

It cant be that bad can it?..I mean I’m on Cubase 5 at the moment and looking to upgrade to 8/artist 8, just doing research into what people are thinking about the products and came across this thread.

I would of thought there would be a huge improvement with regards to the sound engine & interface ect…

I didn’t see this issue (different filter settings in Cubase 7.5 vs. later, so that projects are no longer the same sound) in the “Issues” forum, confirmed or otherwise.

Has Steinberg acknowledged this is a problem?

I haven’t read all 5 pages, apologies if that has been answered already.

FWIW - seems like an easy fix if indeed this is the case - have a toggle in Cubase 8 (or 9, etc.) for, “Cubase 6.5 default filter settings”.

Hello!

Is there anybody in there?

What the (fill in your favourite curse) is going on here! Where is the response from Steinberg on this topic??? :astonished:

I just moved from Cubase 6.5 to Cubase 8.5 (with some hesitation being very happy with 6.5) and wanted to finish a project in it. Everything sounded different, I noticed some distortion and other strange issues. Where does this come from??? It is the same as changing an ingredient in Coca cola and thinking nobody will notice???

I did the null test (same project in mixdowned C6.5 and C8.5)
The result is shocking!! The result is a flanger-like sound, with sudden suppression of random bass notes, clear vocals coming through with an out of phase chorus sound on it, and an overall 10 kHz attenuation of all the tracks coming through clearly. How can this possible??
Please notice that I am talking about the exact same project, all plugins working properly in both C6.5 and C8.5. I am very worried right now and stick to 6.5 until this issue is explained or solved!

Low pass and high pass have changes big time in Cubase 7 or 8.

The filters have evolved :slight_smile:
I rather liked the filters in 7.0, but I prefer the flexibility that came in 8.0
What I don’t understand is that none of the filters use the same Q as in 7.0
Or in other words, that they made one preset that was identical the 7.0 filter.
But things evolve, the EQ changed in was it 6.5 ? to use the StudioEQ algorithms.

I agree, the filters are much better in filtering out, the curves are more steep in my opinion and thus filtering the unwanted frequencies much better.

Ok, it sets me at rest that it is at least pinpointed where this came from. Although it still does not tackle things a phaser-like sound and distortion but as I understand it I will theoretically be able to make the exact same sound if I start a project from scratch in 8.5. The problem only occurs when opening an older project in the new Cubase. That’s a relief, but as a matter of fact it is therefore not backwards compatible I understand. Would be nice of Steinberg at least to mention this for the unsuspecting weakhearted customers! I’m happy there will be always room for improvements :slight_smile:

Hello,

I think the change was mentioned, AFAIK in an ‘Issues and Solutions’ paper, can’t remember exactly where and when, sorry.

At the time of Cubase 7.5 I did discuss with the developer about filters and EQ, if the info can be useful.

About slopes:

Type II reacts to Q-factor and can be adjusted to a very high resonance, Type I reacts to Gain, with a rather subtle resonance. That is for compatibility reasons. Type II was introduced with StudioEQ. Type I is the old Channel-EQ.

For small values of resonance they can be set to be (almost) identical, for example:
Type I Highpass: Gain 0 dB, 1 kHz, Q 1
Type II Highpass: Gain 0 dB, 1 kHz, Q 4.1
<<<

It is worth mentioning that up to 7.5 the filters (Pre section) where 24dB per octave only. Since Pro 8, you have various slopes, the default is 12db per octave.

Thanks Fabio, interesting background information…