Doktor Avalanche
1) Link to quote please ?
KPerry
As someone's already posted, DAWBench is the web's unofficial offical DAW benchmark. I think it has a lot of flaws, but it's the one that is referenced in, sayy, Sound On Sound magazine, and has been discussed here and on places like KvR. I don't know why SONAR is no longer benchmarked there (whether it's down to it not being cross-platform or because the site's owner has had some issues with Cakewalk over the years), but the latest discussion I can find http://dawbench.com/dawbenchdsp-x-scaling.htm - which is very old, admittedly - puts SONAR's performance so far behind everything else's as to be laughable. Mud sticks, unfortunately, and if that's what the stats seem to say, then it's very hard to shake people's belief in them.
Doktor Avalanche
Well as you state very very old. A benchmark of Sonar 7 running XP. That is in a galaxy far far away...
It's time for new benchmarks. If I were Cakewalk marketing I would run their own benchmarks as a comparison, if they are convinced they are good enough (which I suspect they are) I would be contacting journalist, or hire a firm to create independant tests which they can blog about and use in marketing.
Indeed, that would not be a stupid thing to do :-) Still needs work, as you have to compete with what people are used to (like I said, it's like 0-60mph figures for cars...they may be unhelpful, but they're a well-known reference point). But as I said originally, that's a perception that needs to change, so we're agreeing that performance is a perceptual issue, even if not an actual one.
Doktor Avalanche
2) Plugins apart from buggy plugins work perfectly well. Just because Acoustica says so (URL?) does to mean to say it is true. Frankly I doubt it when thousands of plugins work perfectly well..
You've given us the URL to the thread, this is the interesting post within it however:
http://forum.cakewalk.com/FindPost/3236581
If Cakewalk haven't had any contact you can't blame them. Who know maybe afterwards it happened and things got smoothed over. Nobody can assume however there is a problem with Cakewalks code here from this information.
Er...actually, I think you can. There is a definite implication that SONAR bounces/bounced threads between cores which is i) unusual and ii) definitely not good for cache locality and hence performance (that's an unarguable fact of multi-threaded processing). The latter is reflected in the general statistics and reports that show SONAR is/was significantly worse performance-wise than other applications (still, not just SONAR 7 on XP).
Doktor Avalanche
3) Sonar is reliable. It's people's systems that are unreliable. What is Cakewalk going to do gag these people who scream at great heights crying wolf? Admittedly there are still very visible bugs that need addressing (how long must I go on about that) but they aren't stability issues.
KPerry
Well, I find is reliable. But when you hear repeated stories (not on this forum) of SONAR glitching with 2 audio tracks, no plug-ins, or crashing when other DAWs don't, you have to understand that this is what perceptions are. Sure, it might be for good technical reasons (I think the Bakers code more by the book than others, which is why there are more VST incompatabilities than with others who just get VSTs to work like Cubase!), but that's not really something that bothers end users: why spend days troubleshooting something trivial when another application will work with no effort?
Doktor Avalanche
As I've stated, when there are serious stability issues these forums go beserk about the same thing with reproducible steps. All we see right now is stability issues with specific PC configurations. If there are issues with third party VST code Cakewalk cannot be called out for that, bare in mind the vast majority of VST's work fine. I think the actual problem is that third parties are testing their code with cubase but are not with Cakewalk, again no fault of Cakewalks. If the plugin claims to support Cakewalk they should bloody well test it under the platform IMHO. Users should be writing to the third parties and ask them how they do their QA under Cakewalk, is it just a matter of waiting for customer complaints and fixing them?
I'm not referring to these issues: look outside these forums and you'll see general stability/reliability issues that affect users on SONAR and not on other DAWs on the same hardware. It's tough, but the conclusion that gets drawn is that SONAR is finickety/unstable - there was an entire range of audio controller chips that SONAR wouldn't work on a few years back (DICE 2 IIRC) that nothing else had problems with. Why's that?
And, yes, there should be testing with SONAR in addition to Cubase, but that doesn't seem to happen, and we already see a number of plug-ins where SONAR isn't on their compatible list. Why is that?
Doktor Avalanche
BTW the only way is to do it by the book. If Cubase screws up an implementation, and Cakewalk does not, Cakewalk should not be expected to bend their implementation to be like Cubase's buggy code, what Cubase should be doing is fixing their code. It would end up being a spagetti situation for third parties otherwise, i.e. what happens if Cubase then fixes their code to be like Cakewalk's again.
Unfortunately, that's not how the world works. If 9 out of 10 DAWs are written to "behave like Cubase" and yours doesn't, then yours is de facto the incompatible DAW.