Null testing works for me—on straight-up audio tracks with no plugs and no timestretching/tempo-skulldugery of any kind.
I tried one where I had a vocal and a karaoke track. I had done tempo-mapping in 8 to align Cubase with the karaoke. I exported mixdowns with ALL plugins bypassed from both versions. Interestingly, the vocals nulled. But the karaoke only slightly nulled! So time-stretching is happening differently in C8 than 7.5. I’ll have to deeply check settings (time algorithms, etc)… Something to watch out for!
I tried turning off Musical Mode for the karaoke in both Cubase versions—this made for a different-sounding null, but still very partial-null.
So time-stretching is happening differently in C8 than 7.5.
After 4 years of complaining about how bad the elastic algo in previous versions was I’m pleased to see that it’s different…at least shows they are trying. Unfortunately, it’s only different, not better
Yeah, at least they’re working on it!
Their attempts to do time-stretching were so bad at the start (C4-ish?) that I have basically ignored the feature over the years. Recent attempts to use such features have given pleasantly surprising results—so they’ve made headway. But some really odd behaviours remain…
Have you tried repeatedly mixing down and nulling mixes from old Cubase and then C8 alone? Rule out any timing errors that could be happening within the computer?
Earlier, I bounced down audio in 7.5 and 8. I imported both into new projects (7.5 & 8). If I add two audio tracks and import the same source into two tracks (i.e. 7.5 & 7.5 or 8 & 8) and invert the phase on one, then it’s null. But when I have one 7.5 and one 8 (same audio obviously) and invert, then sounds bleed through to varying degrees.
The results from the other PC systems (Windows 7 64 bit) were, as I mentioned before, different but not as expected.
I was chatting earlier with someone who’s just upgrading to 8. He uses a mix of DAWs in his work. He has come across this with one project he has just started work on. It was mixed in 7.5 but in 8 he said it sounded awful and was problematic with odd phase problems. His client noticed the issue too. However, he’s just gone back to working with 7.5 for now as time is money etc etc… Back in 7.5, everything is fine again for him. I was hoping for more info from him but he’s not one to get involved in the details of problems and would rather find a fast solution that works. I’m coming around to that myself.
I do not know what is going on with 8. It’s bewildering in a way. I’m getting horrible performance issues too with C8P, ones that I just never ever had with 7.5. I know I am not alone in this either, but by the same token, I know plenty have had no problems and have seen performance improvements. But that inconsistency troubles me.
I know that the core has been re-engineered in 8, but without having access to the source code, we have no way of knowing what’s going on under-the-hood.
They do null, time to look at what is different in the preferences of the two versions.
Something simple as not having direct monitoring engaged could lead to signals creeping in outside of cubase.
Cricky that really should not be happening, not sure why it is - something is a variable.
I just don’t believe it’s CP8 - there’s no reason CP8 can’t do maths the same as any version before it
or any other DAW for that matter - it’s basic cook book stuff.
There’s something amiss, and it could be a bug in the code, given that the Cubase engine has been re-engineered for 8. I’m not saying that’s what it is, just that it is a possibility.
All I add to it is that I now know a few people who have problems with 8 (and others who do not). Having seen the issue on someone else’s system today was reassuring for my gear.
With regards to preferences, I’ve recently reinstalled everything from Windows up to try to get to the bottom of a performance issue I have in 8 that never existed for me in 7.5. As a result, both 7.5 and 8 were using the defaults.
I’ll mention again that I’ve tried 3 audio interfaces and it made no difference.
I had an email today from a guy who had had the same performance issue that he’s not been able to resolve either. So some people are having problems.
In 7.5, I could quite happily work on complex projects at 24/96 with no major performance problems. I’m not getting that in 8, despite different settings and interfaces.
For now, I’m going to keep my fingers in version 8 in the hope that the next update will resolve these issues. But for my actual projects and work, there’s no way I could use it with the current problems I’m experiencing. It was bliss this evening to do some work in 7.5 with no issues cropping up.
The mix I tried earlier, it was way bassier, tubby, even, coming from C8 as compared to C7.5.
I have spent a couple hours hunting down what could possibly make them Sound so different. The result? EQ.
C8 is processing C7.5.'s EQ differently. They’ve changed it somehow.
I’ve got to do one more pass to figure out if it’s the EQ section or the pass filters in the Pre.
Anyone else want to try this? Take a big mix, remove all inserts. Have just EQ and Pre filters active. See if there’s a difference between your old-Cubase and C8
I’m not in my studio now, and not sure if it helps, but in my case sound was bleeding through in the bass range and in other quite narrow ranges in the mids and high mids.
And done! It’s the Pre filters. Especially the hipass.
I’ve made a dumb, simple loop project with just three stereo audio tracks. No inserts. No processing. Straight bounced mixes of this project from both versions, without processing, will null each other perfectly.
Then…
I did a bounce in C7, then a bounce from C8—with the Pre hipass filter set at 60Hz. C8 was at its default 12dB/octave rolloff.
I then bounced out of C8 with 24; 36; and 48dB rolloffs. The files are named accordingly (e.g., “cubase 8 60Hz with 24db/octave rolloff” = “DUMB LOOP TEST-HP-60-24-C8”.
Here are the results:
The filters in C8 are totally different than C7. That’s fine… things change, but I’d sure like to know about it. Any mixes done in C7.5 are now dead in C8 (even if it did work properly)—they’d take hours of careful work to get back to the mix as it was in C7.5.
Unless, of course, I’m missing something. I’d love to be wrong and know that I can use C8 to open old comps.
Bounced mixes from C8, one each with hi- or lopass filters, 12dB/octave and 24dB/octave null each other out completely. The other rolloffs did not null.
Bounce a mix with nothing else but Pre lo/hipass filters at a certain centre frequency at 12dB rolloff
Repeat at 24dB rolloff
Null test the results
i.e., there is no difference between 12dB and 24dB rolloffs in C8’s Pre section filters!?
As someone who does not work with Cubase for a living, just a hobby. Isn’t it the first law of music production to not change anything of your system with ongoing projects. The second law being to test and test before changing anything of your setup.
NEVER use a X.0.0 version in your paid work. You might be in for some unpleasant surprises. Bad business IMO.
If you are running a studio as a business, you should at least have a couple test computers running. In the long run, it can save you both a lot of money, and a lot of headache.
Computer, Software and plugin testing should be done in your spare time (or planned maintenance time).
And it is also important to make room for studio maintenance time. Don’t change vital studio parts (computer, DAW software, etc) in the middle of a project.
