Appeal for Optimization

[Timmy S]

Ok I know this has been covered a lot, but lets make this be the official thread beseeching MOTU to optimize DP at least once in OS X. We do not need any new features PERIOD. My dual 2 Ghz G5 feels the equivalent of what my single porcessor G3 performed in DP3. I get about the same track counts but can run slightly more plug ins but not many more. I just did a session with an industry pro and he likes DP's features but of course uses pro tools for vocation. But he was extremely frusturated by the save tims and other basic operations.

Is there a way that certain aspects of DP could be disabled in the preferences in the future ? Most people probably only employ 25% of the immense feature list. if motu doesn't spend some time making there DAW the same speed that others operate at I fear there is going to be a prematurely severed future for DP. Come on Motu give us veritable CPU briskness, we deserve this. I don't need to be able to slice up a pitch corrected vocal and turn it into a auto mapped techno beat, I need to save and record and mix, with out perpetual CPU spiking and drop outs. Motu has got to stop denying the validity of the manifold claims to this end.
Please let me know what you think.

thanks,
Tim

[Timeline]

I posted this in feature requests thread before you started this thread:


Timeline Posted: 02 Sep 2005 01:57 am Post subject:
A few more for workflow and CPU:

1. I would like to see all special DP features be normally OFF with a feature I would call "AUTO OFF" as a preference button next to each of the special features in the preferences window.

"AUTO OFF" SPEC:

Special features should remain off until you try to use it, then a placard should appear to ask if you would like to turn it on and you would answer yes or no. After you close DP it would revert to off again.

I think this would be a well accepted preference that would save a bunch of headaches when large sessions are in play.

Also again, the DCPUM would also gain this button.

The most important DP features that need "Auto Off":
* Beat Detection/Tempo
* Dynamic CPU Managment
* Background Processing
* Pitch

"AUTO OFF", It's about time!
_________________

[chrispick]

Ok I know this has been covered a lot, but lets make this be the official thread beseeching MOTU to optimize DP at least once in OS X. We do not need any new features PERIOD.

I happily second any effort to better optimize DP for basic DAW tasks, but I also put its broader, newer feature sets to use all the time. I welcome the advances and use them to generate money. As of late, I've had to commit all sorts of cross-hybrid musical ideas to track, particularly the now-ubiquitous orchestral agro-techno vibe. So, although I commiserate with your frustration, don't throw the baby out with the bath water.

[Jim]

I think the original post is right on, and I put in my vote for no new features until DP is optimized for OSX... I'd even pay for that and be willing to call it an "upgrade."

[Shooshie]

I think the original post is right on, and I put in my vote for no new features until DP is optimized for OSX... I'd even pay for that and be willing to call it an "upgrade."

Ditto here. Though on some quirky, perverse level, I feel that they almost OWE it to us. How radical is that?


Shooshie

[Rush909]

I second tha motion... Atleast fix the serious ones...

1. Many soundbites slowing DP down... this started in 4.5... It's really annoying to have to keep clearing the soundbits list every few hours when working on a heavy project. and the problem is that eventually when working on a big project, you get to a point where you can't clear the soundbites anymore cause you are using them!

2. DCPUM... have not upgraded to 4.52 where this would affect me, but it does not sound like it added as much as it has taken away... Atleast if you don't make it defeatable, then atleast do it like logic... in logic, when you bypass a plugin, it releases CPU... That way I can control when I want a plugin to release CPU and when I want it to stay put... That is the best of both worlds. In heavy passages, I just bypass a bunch of non-used plugins which will give more headroom for the passage to play properly..

3. ok this is not a bug, but need attention: Freeze function. it's almost useless... Make the thing NOT realtime and it will open up DP wide open, and will make people forget about how DP is not as optomized for speed as other apps, since more people will start freezing because it will be so convenient...

r.

[Jim]

Ditto here. Though on some quirky, perverse level, I feel that they almost OWE it to us. How radical is that?


Shooshie

I agree with your agreement, Shooshie. An optimization "should" have been done on the first OSX build, really, and we paid for that "upgrade" which wasn't really an upgrade, but a port to the new OS. DP actually ran more efficiently and stably (Stably? Yes, that's a word... I looked it up.) on OS9, and still does for some of us here on the list.

The problem with trying to browbeat MOTU into doing an optimization for free was already mentioned... they need cash. But it seems to me that if they're just doing patchwork jury-rigged workarounds on old code with these new features, eventually the crows are going to come home to roost, and the straw will break the camel's back, and add your own hackneyed cliché about how the sheisa's going to hit the fan.

So, why not go ahead and get the optimization done?

It defies credibility that after, what two years of DP on OSX that it still runs less efficiently than DP3 did under OS9. And they still haven't figured out how to bring patch names into DP from Unisyn as well as it was done under 9. Tsk, tsk.

So, my offer to pay for an optimization serves as evidence of my penuriousness ... I would assume it'd be cheaper than moving to Logic or PT. MOTU has my loyalty, for now, but they need to keep earning it. It's not a birthright.

[sdfalk]

It defies credibility that after, what two years of DP on OSX that it still runs less efficiently than DP3 did under OS9. And they still haven't figured out how to bring patch names into DP from Unisyn as well as it was done under 9. Tsk, tsk.

Having used Logic in OS9 (and now owning Logic Express in OSX)
Logic isn't as efficient running in OSX either.
Some of that Tsk tsk needs to be reserved for Apple as well.
Mind you we are talking about an entire OS and not just an app.
I've used OSX since I got the Beta in the mail.
It's gone from slower then hell to just..slow.
The Finder can still be brought to a screeching halt pretty easily.
Scrolling (in any app) can still be clunky.
OSX still needs work to improve the experience of using the apps
within it.
It's not just a DP problem (IMHO)

[Shooshie]

Well, truth be known, I'd pay a full upgrade fee for a version 4.63 that has no extra features except total reliability and optimized speed. So far, the speed increases and stability in DP for OSX have all been of the variety that emphasizes patience. We learn to readjust our expectations and work habits. The whole idea of OSX was that Core Audio was going to be blindingly fast with near-zero latency. I'd say that OS X has not lived up to its promise there.

Funny thing, I pulled out some old files from 1984 the other day. Yes, you read that right. 1984. They were in ThinkTank format, an old outliner. The original outliner, for that matter. I opened them in MORE, which was the evolved version of ThinkTank. It was dated 1986. Amazingly, it ran in Classic. It was blindingly fast. It did what it is supposed to do, and did it well--with a little coaxing here and there. And it took up less than 400k of disk space, and used about 650k of RAM. What went with the days of clean, concise code? I remember interviews with Bill Atkinson where he was proud that he got an entire Hypercard routine to run in the 4k cache, or maybe it was 24k, or something ridiculously small like that. Programmers used to stay up nights on end shrinking their code to absurdly tiny algorithms that were perfect, like a Bach Fugue. Why not now?

I once worked for a software house that made databases in the old Mac Plus /MacSE days. A particular search feature of our software was so slow that the operators (our customers) would have to set it up the night before, then leave it to run for a day to get the results. I went digging in our code and found the reason. There were 87 pages of nested "if-then" clauses in one single operation. Every item had to pass through that whole guantlet of superfluous code. So, I asked our NEW chief programmer, a woman with mainframe experience, if she could reduce that to a few simple lines. She said yes, and the new algorithm did the same job in about 5 minutes.

Pure and simple, it's code problems. We need some ingenious coders at Mark of the Unicorn who are willing to go through it and reduce it. OF course a big problem is that there are so many libraries all interdependent. Change a line or two here, and it might break it somewhere else. It's a job no coder wants. So, yeah, we'll have to pay for it.

Shooshie

[eriknorlander]

I fully completely totally 100% thumbs-in-the-air whoo-hoo hollering amen brother give-me-some-love agree with this thread.

I started with Performer in the late 80s on a Mac Plus with 1 MB of RAM. I bought every update for Performer, and I dragged my feet with "Digital Performer" until the 2408 came out. I used that with a G3 300MHz machine and then bought every update of DP that has come out since.

Now I'm up to a G5 dual 2GHz machine with 4 GB of RAM (more RAM, btw, than the size of the HD on my G3!), and while it is faster, it is not blindingly faster. I would expect the difference to be like that of a WWI biplane and an SR-71 (or some other silly simile). But it's not. The new features and possibly even the newer Mac OSs seem to slow things down, only requiring that faster CPU and more RAM be used to keep up or make things incrementally faster.

The save time, for example. Geez. I often work in projects with multiple chunks, and I have to choose carefully when I save during a tracking session. I don't want the artist looking at me going, "Okay, let's go! Let's go!". I would pay $99 for a faster save time alone.

Anyway, yes, MOTU should optimize, and we should pay for it. I don't think they make any money by posting free updates. While I was startled (and grateful) for the free 4.6 upgrade, I would have paid for it. They are a business, not a NPO, and they should be paid for their work. We can all complain about what they should have done better or differently (and we do), but in the end, they have "finished" their product each time a new saleable version is released, and as long as it doesn't have gratuitous defects (like exploding Pinto gas tanks ... there I am with another simile), that's it. Take it or leave it. So we should be willing to pay for upgrades and optimizations, and we should let them know it. I see some MOTU people posting here, so clearly they read what we write. That's very cool.

$99 for a faster, more optimized DP with the ability to disable features I don't use? I'm in. Right now. Credit card at the ready. $199? Yep. Still there.

Best regards,
Erik

[RCory]

Yes.. me too...

I'd pay for an upgrade such as this...

C

[Timmy S]

well we should all consign emails to motu about this and let the know we need this. i will send them a missive right now. i hope you'll do the same so we may actually have some hope of doing some professionally paced work.

[sdfalk]

I've sent my emails to Motu (as well as Apple) long ago

[DrDespair]

Yeah, software.

It‘s the same everywhere you look.
I‘m very pleased to have a OS9.2.2 G4 Macintosh. Ever opened FreeHand 5.0 on such a thing? 1 second, you hardly notice it.

But here you are as a company, confronted with two groups of customers:
the big majority that is always comparing bells and whistles of logic/cubase/you-name-it, and never bother for real performance issues, because they use it as an expensive••¦ TOY!

and then you have the pro‘s that use TOOLS (thats why they call it Protools, I guess)

the majority pays the bills.

I‘d say I worked much faster on my Yamaha AW4416. But absolutely no Backup-Possiblilty, and a fan that‘s LOUD made me switch over to a MAC-based system. (the G5 sits in another room)

How about Digital Performer Basic.
tracks, punch in out, some markers, mixing board, basic automation.
zero latency.

and then buy the bits and pieces you think you need additionally.

DrDespair

[BKK-OZ]

If by optimization we all mean not just more efficient, but also more solid, I would gladly pay. I can't remember a single project where I have used, say, 1/10 of what DP can do (feature-wise), but conversely, in every project I have some sort of crash, or problem that I have to work around.

Some of the crashes I have I know (I think!) are caused by plugs, and that's not neccessarily something to blame MOTU for (tho, I think they should share in the blame), but many crashes come direct from DP.

Just tonite I had another - I have been working on my latest masterpiece (it really is a masterpiece, trust me!). Because I know that Guru can be flakey and cause crashes that end up locking me out of files (can't even load the sequences) I have taken to saving my MIDI as clippings, so I can at least get the MIDI data back. Well, I went to save a clipping, and as if DP were saying to me, "you've had it good for about 4 days without a crash, time to remind you how tricky all this stuff is!", the whole thing went up in smoke.

The spinning beach ball from hell haunted me until I got the "The Application formerly known as DP has unexpectedly quit..." dialogue box coming up.

Argh.

I wasn't doing anything bad (at least with my computer!), this stuff just shouldn't happen any more.

Time for the bloat to end, lets just get something that is rock-solid.