Grr why are older versions of software so often better??

So I am running Cubase 6.5 on Win7 Pro x64 on an Intel corei7 3770k with 16gb RAM and SSD drive, etc (upgraded from cubase 4 on win xp x32 on an intel core 2 duo E8400 and 4gb RAM). M-Audio Delta 44 PCI card common to both systems. I have 128 samples buffer size, windows optimised for background services, multi-processor enabled in cubase, etc. I was really hoping to see massive performance benefits with ASIO headroom and so on. But being optimistic I’d say it’s barely better in any respect and in some respects it’s just plain worse. For example:

I have a new project with two instances of Halion SE running and just midi files. I’m using 7 patches in instance 1 and 1 patch in instance 2, with some of Halion’s delay and other effects on some of the patches in instance 1. I have a couple of other Cubase effects such as EQ, compressor etc on the output buss. With just that my ASIO performance indicator is already hitting around 40% - but my quad core chip is using only about 12% of it’s capacity and system memory about 18%.

So when I play the track the delays run in time (they are synched to the tempo) but when I export the mixdown the delays end up being massively OUT OF SYNCH. If I do a REAL TIME export the delays are in synch. But the track is 60 mins long. And I have several to do like this.

Do I really have to sit and wait for an hour to do a mixdown just to have the delays processed in time?

I’m using 7 patches in instance 1 and 1 patch in instance 2,

Aloha S,

Just wondering if one (or more) of those patches is called: "Sapphire?
That one seems for me to ‘sop up a lotta juice’.
I now as much as possible stay away from it.

Do I really have to sit and wait for an hour to do a mixdown just to have the delays processed in time

Would it work (or at least save a lil time) to first render those
Instruments tracks to audio and then do a mix down?
(Just spitballing here).

Hi curteye

That’s what I had to do - create loops of composited midi tracks with reverb and delay tails, re-import them, cut them up, re-export the full hour-long mix. Great productivity boost eh? People manage to come up with workarounds when software doesn’t perform properly (another example would be the appalling file browsing/new project issue as mentioned in my other recent post!) but the question is, when they’re shelling out large wads of hard-earned cash on these products, should they have to?

It’s just another incident in the long list of incidents which bring home the truth of the ‘upgrade myth’ - software / hardware - it’s so often just a new-grade. We’ve got used to paying for beta software and hoping for bug fixes to come out just before the next unfinished new-grade is released… Bill Gates is the main culprit of course, but he’s by no means a lone shark in these waters…

I have a completely different experience though I didn’t upgrade Cubase only my DAW.

I was coming from a Core2Quad q6600 and went to an i7 3820, 16 Gb, SSD (boot) disk.

A song that would sit at 50 % now comes in at 12 %. And the best thing: I can play the biggest patches of Diva in the “divine” mode at full voice-count and come in at around 30%. My old PC wouldn’t even play one voice in that mode and with the Jupiter 6 modules (which are the most CPU demanding ones in Diva).

I don’t need to bounce anything anymore, I don’t need to switch to higher ASIO buffers for mixdown, Cubase starts from the SSD in a couple of seconds and I’m a pretty happy camper. 500 Euros well spent for that DAW upgrade (I kept all other hardware and the case)

Older versions are “better” because:
a: New features mean new resource paths and some new bugs and
b: You missed a new setting/s.

I don’t think it is a general rule at all. I find Cubase 6.5 to be better than any previous version. With Windows from Windows XP to Vista was definitely a backward step but vista to windows 7 was much better (it looks like Windows 8 will not be so good though!). So I guess it depends although there are certainly many companies that add features first to sell a new product rather than tackle bugs. I don’t think that is true of Steinberg though.

There are better features of course, that’s the whole reason you’re enticed to upgrade. Yes it takes some getting used to new things - I’m not anti-change in any respect but what really gets my goat is when the very basics are missed, or things that are just common conventions are pointlessly deviated from - such as file > new calling a tried and tested folder browser dialog class that lets you actually see all the folders on your hard drive, instead of trying and failing to reinvent the wheel and ultimately putting a load of pointless obstacles in your path that you have to work around just to start a project!

One of the primary functions of a DAW is to be able to playback and mixdown mixes accurately (i.e. what you hear on playback ought to be what you get in mixdown right? or is that too much to expect?) and if mine’s failing to do that, using only Steinberg products on a hugely capable system with an experienced driver, then it’s a pretty basic failure don’t you agree?

Asrock Z77 Extreme6, Core i7-3770K CPU @ 3.50/3.90 GHz, 16gb 1333 MHz Kingston RAM, SATA Sandisk 240gb SSD primary drive, several TBs of SATA HDDs, Win7 Pro x64

Yes something is not right if that is happening. Mine is fine but it is the nature of PCs that they vary so much that something works fine on one and not on another. But I agree you should expect export to work correctly if everything else is set up fine.

Yeah, that’s a failure, but unfortunately (for you) on your particular system. I’m enjoying the best version of Cubase yet, and would never return to any of the previous systems (though I would like an afternoon with my old Atari ST…). Are you running the latest drivers for the M-Audio? Must be something

OK, someone do me a favour then an emulate what I’ve done - start a new project, fire up an instance of Halion SE, load 8 patches into 8 slots, set up 8 midi tracks and bang in some midi parts for each track. Pick some floaty soundscape patches that have their own delays and get them running around each other a bit. Then add Multi Delay to FX1 and set up 1/2 synch in stereo mode, some L/R delay, 80% feedback, and run a few of the patches thorugh it. Add Stereo Pan to FX2 and run some of the patches through that. Get it so you have some delays working around each other but so that they are pleasing to the tempo. Then do a mix down and see if the delays stay synched the same as in playback. And if not, see if a real-time mix down solves it.

I’d be interested to know if it’s just me…

Older programs run better on modern computers?

As far as the delay print to your file… Try a different plugin. Try changing buffer size. I dunno. Try a host of things to narrow down what is going on.

If it does it with a different plugin, for instance, or even maybe a few different plugins, you know it is not the plugin.