Things I hate about... - Cubase 7

[Tp3]

So yes - it's a WYSIWYG readout comparison.

In that case the comparison is dead WRONG (in my very humble opinion).

Sorry, but these "benchmark tests" use too many demo versions for my liking. Why can#t this test be performed with freeware or host internal stuff.

Well, I don't want to be a stick in the butt but R. ... you are SOOOO mad that you can't let yourself test the FACTS.

And the FACTS are that DAWbench is using a custom-made ReaXcomp SPECIFICALLY developed for this test, under supervision from Vin (it is bundled with the DAWbech zipped file).

Did you "optimize" your system (OS) in some why ? with every version it seems that Cubase "demands" optimization-free system.

Just me 0.02...

[Compyfox]

So yes - it's a WYSIWYG readout comparison.

In that case the comparison is dead WRONG (in my very humble opinion).

Care to elaborate on that?

Well, I don't want to be a stick in the butt but R. ... you are SOOOO mad that you can't let yourself test the FACTS.

I have tested enough already, don't you think?
Considering 6 OS nukes, several tests in between, various drivers, even different rigs.

And the FACTS are that DAWbench is using a custom-made ReaXcomp SPECIFICALLY developed for this test, under supervision from Vin (it is bundled with the DAWbech zipped file).

I only see two Benchmarks on DAWbench these days. The refined with ton of "demo" plugins for processing performance, and the one with Kontakt 4 for MIDI perfomance.

Am I missing something?

Did you "optimize" your system (OS) in some why ? with every version it seems that Cubase "demands" optimization-free system.

Let's me rethink what I did up until this point...

- OS nuke, several times, check
- "old" drivers (read: absolute minimal) up until most recent for RME and GPU, check
- no hybernation, testing if page file has an influence, general OS treatment, AERO on/off, SP1, all maintenance fixes, check
- CPU parking fix, check
- Power Management at max so that nothing turns off, check
- no useless drivers/tools/firewall/etc, check
- general OS treatment towards "audio applications", check
- the Steinberg internal "switch" to "optimize performance", check - resulted in a slowed down rig
- Overclocking the CPU of about 900MHz, check
- testing if EIST/SpeedStep and C1E (CPU power managment) has any influence, also check
- Cubase in various flavors (C6/C6.5, C7.02 to C7.06, 32bit/64bit) - check, sometimes 32bit performs better, sometimes 64bit (depending on the plugin, especially if they're available in both formats)


Did I miss anything, especially something I have mentioned before? Dunno, it's been so much that I actually did in order to try to get things running properly.

[lotus2035]

OK I bit the bullet and bought the upgrade from 7.0 to 7.5 which cost next to nothing.

you must be a rich guy when €50 is next to nothing for you.

I'm far from being rich but I don't drink or smoke and I don't have any kids or outstanding debts. It's all relative.

[ZINO]

OK I bit the bullet and bought the upgrade from 7.0 to 7.5 which cost next to nothing.

you must be a rich guy when €50 is next to nothing for you.

I'm far from being rich but I don't drink or smoke and I don't have any kids or outstanding debts. It's all relative.

The problem is not the 50 Euros "per se". The problem is the new practice that is solidified by Steinberg to charge for point releases along with the already high priced major upgrades. And these paid point releases are unleashed even when the product is still in a semi-beta quality stage when even the GUI is redesigned every month. Sad...

[murnau]

yep, that was also a principal question for me. i have enough money left (it's a shame) i bought this month for 800 eur all in all but i'm not willing to gave steinberg my money anymore. instead i bought a number of high level plugins like slate and kush and k9 etc.

[Tp3]

Care to elaborate on that?

I can't quote directly, I only vividly remember one of the mods stating that the new meter is taking into account a lot of things besides ASIO load, and that the two meter can ABSOLUTELY NOT relate one to another. I remember this is the one and only reply from Steinberg rep I had seen on that matter. it was last December (2012).

I have tested enough already, don't you think?
Considering 6 OS nukes, several tests in between, various drivers, even different rigs.

Dunno, I'm not sure the methodology is right. of course, my personal opinion.

I only see two Benchmarks on DAWbench these days. The refined with ton of "demo" plugins for processing performance, and the one with Kontakt 4 for MIDI perfomance.

Am I missing something?

Yes.

Download ---> http://www.dawbench.com/downloads/dawbe ... 012-r3.zip

In there you'll find one folder and one file that of interest to you :

* Folder : Cubase (+the "Audio" folder)

* File : reaxcomp-standalone-dawbench edition.zip

Copy the folder and the audio to temp location. copy ReaXcomp to your vstPlugins.

Run either "DAWbench-DSP-C6-RXC.cpr" or "DAWbench-DSP-C6-RXC-EXT.cpr" (the latter double the cpu drain it imposes. it is intended for modern day DAWs).

Check your results against This and That (roughly, of course. but you should have an idea as to what to expect from your system and if you are in the ballpark).

Did you "optimize" your system (OS) in some why ? with every version it seems that Cubase "demands" optimization-free system.

Let's me rethink what I did up until this point...

- OS nuke, several times, check
- "old" drivers (read: absolute minimal) up until most recent for RME and GPU, check
- no hybernation, testing if page file has an influence, general OS treatment, AERO on/off, SP1, all maintenance fixes, check
- CPU parking fix, check
- Power Manamgent at max so that nothing turns off, check
- no useless drivers/tools/firewall/etc, check
- general OS treatment towards "audio applications", check
- the Steinberg internal "switch" to "optimize performance", check - resulted in a slowed down rig
- Overclocking the CPU of about 900MHz, check
- testing if EIST/SpeedStep and C1E (CPU power managment) has any influence, also check
- Cubase in various flavors (C6/C6.5, C7.02 to C7.06, 32bit/64bit) - check

Did I miss anything, especially something I have mentioned before? Dunno, it's been so much that I actually did in order to try to get things running properly.

Well, as I've said, from version 6 (and Win7)it seems that Steinberg "NEEDS" a "clean" system.
no special "treatments". they seem to only F.Uck things up.

Maybe the OC also F.Ucked things... who knows. you shouldn't rule out ANYTHING (my personal guess is that it is down to two things : your OS and your expectations).

[Compyfox]

Well the PDF I posted earlier clearly shows that both the OS and the OC didn't mess anything up - most tests were performed post(!) OCing and OS reinstalling. The tests with 7.04 were also performed on a clogged system (an old image from pre OCing), and then rechecked on a "clean install" (post OC).

I didn't need to change any value I've encountered up until this point - which should be one indication!



Of course the Steinberg reps say, that the "new system" can't be compared with the "old" engine. I do understand that to a certain point. But if I have a 16 channel WAV project at 48khz, with only 3-4 inserts each (and those being low profile ones - meaning CPU optimited, not even internally OS'd until rendering) has about 30-40% more load in C7 compared to C6 - I do consider this a severe issue.


Thanks for pointing out that there is "another" test hidden in the current DAWbench project. It's not mentioned on their page. Maybe I'll take another(!, sheesh... ) dive at this and compare it with ScanPro Audio's results - now that my CPU clock comes close to their Sandy Bridge listings (even though I'm still on a Bloomfield!). I trust these people - they helped me during OCing, they recommended me my rig. And they were the first to shake heads on these results.

But only if I am in the mood still. Not today though...

[Tp3]

Well, DAWbench is the "standard" bench, everybody should use it (IMO).

It usually "detects" problems with your rig and/or you card drivers. Vin is one smart MF. he has a new bench in the works (with a new reaXcomp, but they will appear in 2014.[/list]

[SODDI]

You can test and test and test. 3 things are guaranteed:

1. Cubase 7.0x is not going to run any better on your rig.
2. Steinberg is not going to have any solutions.
3. The Cubase fanboys are still going to insist that's nothing's wrong, that it's all your fault, or Intel's fault, or anyone's fault except Steinberg. After all, it runs OK on their rigs, therefore it must run perfect on EVERYONE'S rig.

[standalone]

The Cubase fanboys are still going to insist that's nothing's wrong, that it's all your fault, or Intel's fault, or anyone's fault except Steinberg. After all, it runs OK on their rigs, therefore it must run perfect on EVERYONE'S rig.

Where do you find such bizarre people? I don't know anyone like this.

[Tp3]

You can test and test and test. 3 things are guaranteed:

1. Cubase 7.0x is not going to run any better on your rig.
2. Steinberg is not going to have any solutions.
3. The Cubase fanboys are still going to insist that's nothing's wrong, that it's all your fault, or Intel's fault, or anyone's fault except Steinberg. After all, it runs OK on their rigs, therefore it must run perfect on EVERYONE'S rig.

Well, I hope I'm not regarded as a "Fanboy" (on the contrary).

The thing is : there's always a solution. ALWYES.

For instance, years ago, I campaigned with another guy FOR MONTHS because of a defected dongle code that made the visual operation sssslllllloowwwww. Steinberg fixed that (it was from 5.0x-->cubase 5.5) but it was aggravating.

Problem is : they can't deal with SOLITARY systems. I can understand that. I'm just suggesting a solution.
As of now, I think that the fastest solution will be to install CLEAN os on another disk, NOT TWEAKING IT AT ALL, and trying to see what happens from there. it ain't pretty, but it's the fastest solution. the alternative is to keep on complaining, KNOWING that the complaints will not be answered. EVER.

Or

Move to another DAW.

The thing is : one has to correlate his expectations and reality. this is not "Fanboy Bible". it is coming from YEARS AND YEARS of experience. anger, aggravation and experience.

Compyfox : PLEASE check DAWbench. this is the most reliable, thought-out, honest to the bone - audio benchmark out there, as of today (and it has been like that - for years).

[mkdr]

I want a healthy headroom.

Do you even know what a headroom is?

If you feel that 100m sprinters should be clocked at the 36,7m mark and judged by how close they run to the time set by King George VI.. then fine. But don't come telling us that Bolt didn't win even though he was the fastest (and doing it in every topic here...).



I have left this thread now for good

So Long, and Thanks for All the Fish

[fandango]

I have left this thread now for good

[SODDI]

You can test and test and test. 3 things are guaranteed:

1. Cubase 7.0x is not going to run any better on your rig.
2. Steinberg is not going to have any solutions.
3. The Cubase fanboys are still going to insist that's nothing's wrong, that it's all your fault, or Intel's fault, or anyone's fault except Steinberg. After all, it runs OK on their rigs, therefore it must run perfect on EVERYONE'S rig.

Well, I hope I'm not regarded as a "Fanboy" (on the contrary).

The thing is : there's always a solution. ALWYES.

For instance, years ago, I campaigned with another guy FOR MONTHS because of a defected dongle code that made the visual operation sssslllllloowwwww. Steinberg fixed that (it was from 5.0x-->cubase 5.5) but it was aggravating.

As of now, I think that the fastest solution will be to install CLEAN os on another disk, NOT TWEAKING IT AT ALL, and trying to see what happens from there. it ain't pretty, but it's the fastest solution. the alternative is to keep on complaining, KNOWING that the complaints will not be answered. EVER.

If you think that wiping your computer clean and re-installing EVERYTHING from the ground up just to get a single program to maybe, kinda, hopefully sorta work properly, wiping 100s of programs that operate properly, is a GOOD solution - there's something wrong there.

It's not just Compyfox, and it's not just here on KVR - the Steinberg forums are filled with the same kind of problems and have been since Cubase 7 dropped (if Helge Vogt hasn't censored them out of existence.) Gearslutz if full of the same issues.

Solutions? I've gotten solutions up the wazoo, including one from the California Steinberg techs that wreaked havoc on my system - and then had Steinberg's Guillame Navarette condescendingly lecture me that what I had to do to compensate for his own techs' idiocy was wrong. (The best solution came from a KVRer who suggested that I change the the Synchrosoft dongle to another USB bus. Worked like a charm, proved my assertion that my MOTU USB MIDI device was functioning flawlessly. But the Steinberg techs didn't seem to know nothing about that.)

[D.H. Miltz]

I have left this thread now for good

This is the second (and last) request that you do this, and stick to it. Please and thank you.