@julian said:

DG, Full boot time for 10 collections on a MacBookPro 2.16 running Windows XP2 (if you can stomach the 1980's graphics!) takes 2 minutes. Same computer running OSX takes 5:45.

Re crashing on a Mac - if I run my audio apps conservatively it will go for weeks without a crash. When i run everything against the end stops and the computer is straining away you can get more frequent crashes. Most of these, due to the stability of OSX, don't affect other active programmes or the "finder" but good practice should dictate a OS restart in case of any unnoticed corruptions.

Julian.

Yes, I thought that you said that all 10 collections were about 2 minutes to scan. That's why I was surprised at Jay's 10 minutes.

I'm glad to hear that Macs don't crash that often. I am always surprised to hear that any computer crashes in normal use, as I think that mine have only crashed once in the last year (over-heating, as I couldn't be bothered to switch the air con on!). It is good to know that both platforms are stable and will continue to give people a choice.

DG

it sounds like some are wondering about the difference of needed time for scanning licenses between the license control center and the Vienna Instruments

well - LCC simply looks which licenses are on your ViennaKey (this is currently a maximum of 10) and displays their names, no check of any library data happens there, no harddisk access, ect

the vienna instrument is cycling through the data-files assigned in your directory manager and compares the needed license to access this data with licenses that are present on your ViennaKey

christian

@cm said:

the vienna instrument is cycling through the data-files assigned in your directory manager and compares the needed license to access this data with licenses that are present on your ViennaKey

But the question remains: during this process there is almost no processor activity and almost no disk activity, so what's actually going on? I still think there must be a possibility to speed this up dramatically... Anyway it's good to hear that VSL are investigating this.

DG-- The 10 minute figure is an approximation and includes the loading of 1.5 GB of VSL VI samples. I will put a clock on it and post the exact amount of time in the next day. But it is a lot longer than I can sit quietly just staring at my monitor [:D] .

Best,
Jay

[/quote="DG"]
Why does your computer crash 2 or 3 times a day? Not to detract from the Syncrosoft issue, but this is not normal, so there could be something wrong with your system.

DG[/quote]

Probably because i'm always really pushing things, with tons of tracks and using plug-ins from many different manufacturers. Plus, the tighter the deadline, the more crashes!

There's nothing wrong with my computer, but as a beta tester for various software programs, I'm aware of the significant number of bugs in the programs we all use everyday (even the release versions), and these types of thing usually are most problematic when you're pushing stuff to the limit. Maybe it's more like a few days with no crashes, and then a few with 2 or 3, whatever, but the point is we shouldn't have to wait 10 minutes for 10 licenses to re-scan (on first boot-up) if we have all those licenses on a single Mac.

@JBacal said:

DG-- The 10 minute figure is an approximation and includes the loading of 1.5 GB of VSL VI samples. I will put a clock on it and post the exact amount of time in the next day. But it is a lot longer than I can sit quietly just staring at my monitor [:D] .

Best,
Jay

Jay, that's still a lot longer than my system. For some reason boot-up is longer now I have more memory, but the Syncrosoft thingy is less than 1 minute and 2.8GB samples takes a further 5 minutes (roughly). It would be useful fro you to split up the sample loading from the Syncrosoft issue, as this time is affected by other (hardware) factors. Do a cold boot and then load an empty project, apart from 1 (empty) Vienna instrument and you should see how long it actually takes.

DG

@garylionelli said:

There's nothing wrong with my computer, but as a beta tester for various software programs, I'm aware of the significant number of bugs in the programs we all use everyday (even the release versions), and these types of thing usually are most problematic when you're pushing stuff to the limit. Maybe it's more like a few days with no crashes, and then a few with 2 or 3, whatever, but the point is we shouldn't have to wait 10 minutes for 10 licenses to re-scan (on first boot-up) if we have all those licenses on a single Mac.

Fair enough, at least you have a reason for your computer's poor performance in this regard!! I only asked because many people seem to have no idea how stable computers ought to be with normal use and genuinely think that it is acceptable for them to crash every day.

FWIW I think that you are very brave chap; I couldn't cope with crashes in the magnitude that you're having, as it would drive me even more nuts than I normally am [[:|]]

DG

@paul lm said:

yeah, thanks. please make this a priority. on the soundtrack i'm working on at the moment, it takes so long to load up in between q's - many q's can take 4 or 5 minutes to open, if i have many VI's in the sequence. (Digital performer/dual 2.5 g5), and it's really embarrassing when you've got a director in the room having to wait for it to load up. With ilok, it all loads up instantaneously. This isn't to detract from the VSI itself, though, for which you guys are to be congratulated - major achievement. But if you're going to spend 6500 pounds on a product, like we've all just done, this does leave a bit of a bad taste in the mouth [8-)]

The time taken to load between cues is governed by the amount of sample headers that are required to be loaded into RAM. This currently is a little slower - but not much - than sample load with EXS 24 in Logic.

This thread has been discussing a different issue that is not sample load related but concerns the initialisation time of Vienna Instruments at first boot up PRIOR to any sample load.

You will obtain the fastest cue to cue load times by employing fast hard drives, a fast computer processor and plenty of RAM.

I would hope that there will be optomizations in the future that speed up the RAM loading time. My unscientific calculations give (on my system - G5 dual 2.5/6.5 GB of RAM SATA 2 Raid drives) sample load speeds in the region of 10-15MB/second - this is from a drive set that has sustained transfer speeds of 200MB/second. So sample load on my system is only about a tenth of the systems potential.

I've no idea what is causing the bottlenecks - whether is hardware based or software, but I'm sure there is room for improvement with current technology without having to wait for yet to be released faster processors and drives.

Julian

julian, without hijacking this thread, i can't see my system being that different to yorus? dual 2.5 g5, 6 gigs of ram, the only difference being that my VSI is installed onto a dedicated firewire 800 drive. will this make that much difference compared to your SATA drives? rephrase: what IS a SATA drive [:D] should i swap?

Hi Paul,

A SATA drive is a relatively new disc interface. All drives connected to a SATA bus have their own connections directly back to the computer. Currently the maximum THEORETICAL transfer rate with SATA 2 is 3000Mb/second (375MB/sec) compared with 800Mb/second (100MB/sec) for firewire 800. I use 4 fast drives raided together and they transfer large files at about 1600Mb/second (200MB/second) in a Raid Zero configuration.

In the real world and certainly with Vienna Instruments there is no difference in initial boot times between my SATA drives and Firewire 800 drives. In sample load times the SATA is quicker but only by a few percent. The SATA will stream more samples (greater number of voices) than the firewire. These statements are based on a 4 drive SATA raid - a single SATA drive will not really outperform a single firewire drive of similar type (400 or 800) as the drive will be slower than the interface speed.

I use firewire audio interfaces so the net result of a SATA drive for Vienna Instruments is good for my system as the load is spread and certainly for maximum audio track count SATA (in a RAID configuration) wins.

With a 4 disc SATA raid the chance of failure is 400% greater than a single drive so back-ups are the order of the day.

I was also pointing out in my reply that you stated concerns about cue to cue load times - whereas this discussion was focusing on a different issue - Vienna Instrument boot times.

Julian

"the vienna instrument is cycling through the data-files assigned in your directory manager and compares the needed license to access this data with licenses that are present on your ViennaKey."

Thanks, Christian. My technical understanding is limited. I'm not sure why the Vienna key needs to match such things on a per sample or per instrument basis. What you describe sounds more like librarian work -- the verifying of location of data on the disk. Perhaps "Scanning Licenses" means "Matching Licenses" to that data. I thought it just meant checking to see if the licenses were there -- like the XS key in Logic.

No doubt in the end, this is a security mechanism. I'd hate to think it's a simple librarian task that (as another poster mentioned) could be saved in a preference file. (And perhaps a preference file save would make it more "hack-able.")

Thanks for the response.

update-- I put a clock on loading a Sonar 5 project with 18 optimized VSL VI's (15 if these were disconnected) in it containing roughly 600mb of samples after a cold boot of my PC computer.

It took 9 minutes and 25 secs.

It will try to isolate the sychrosoft portion of that time by loading a project with empty instances and post the results later.

Best,
Jay

If the launch time doesn't change much with SATA RAID then I'd imagine the bottleneck is the Vienna Key itself... makes sense. USB is pretty damn slow, after all.

I must admit that the first-launch time of the VIs is ridiculously long... and I'm only running 4 sets!

J.