64 Bit Pro Tools? RTAS Working on ALL Cores?

[O.G. Killa]

Originally Posted by Bentley Ferrari

O.G. Killa-



You make some good points, but this is not one of them, at least not for a growing number of users. Check out a number of recent and on-going threads in this forum re: RTAS limitations/problems on HD systems. RTAS usually works better on LE systems, often with no latency to speak of.



Now-- your point would be a little better taken if TDM plugins were still being developed like they were 5 or 6 years ago, but they're not! The AIR line of plugins is a case-in-point. That's huge.

So you're faced with a growing number of desirable plugins that can't be used effectively with an HD system, or at least not as optimally as they can be used with an LE system. At the same time you have substantial leaps in native power happening every 18 months on average to run LE and MP at lower latencies with more track counts and plugin instances. HD is not as much of a clear winner as it was when a lot of us first bought our systems. It depends on how you use your system, of course, but IMO it's obvious the balance is shifting.



Which brings us back to the original intent of this thread: What is Digi planning to do about it?


I still think they need to do what other tech companies do and give us something reasonable to look forward to.

I mean, c'mon....It's not like anyone's asking for schematics or source code here.

You make some very good points too. AIR plugins only being RTAS I think is partly a situation of circumstance. AIR plugins aren't made at Digi's headquarters. They bought a company in Germany that made VI's. Since VI's are RTAS, and this company was already experienced at making Native VI's, they continued with that. All of the new AIR plugins are the EXACT same effects that are within the AIR instruments, just stripped out into their own plugins. I think that's where the "matter of circumstance" kind of happened. It's not like AIR designed these effects from the ground up to be stand alone and purposely made them RTAS only. They simply exported the code they already wrote into Hybrid, Structure, Velvet, etc and put them into their own plugins and nobody spent the time yet to port them over to TDM.

But that still doesn't negate your very good point... they won't work as "well" in HD. I don't know if Digi is planning on porting them to TDM, but we might see that in a later release. *shrug* who knows?!

I wouldn't say that TDM plugin development has stopped. Companies are still making TDM versions...it's just that, from what I understand, the developer kit from digi makes it easier for developer to write the RTAS version first, then port it into TDM. I don't write software, so forgive me if I'm wrong about that. But in the past when complaining to 3rd party plugin makers about "when will it be TDM" that is what they have said... and they follow it with, "but it's coming. The only company that comes to mind that doesn't make TDM versions of their plugins is Izotope. But all the other ones I can think of that make RTAS plugins, also have TDM versions (except for VI's). And there are still some really great TDM plugins that Digi hasn't ported (and seems to have no plan to port) over to RTAS. Revibe being one of them.

BUT!!! I think my point is a still a valid one about TDM plugins vs the real hardware... I think part of the discrepancy is in the target markets for each product. While Peppertree calls it a "club", there are features in HD and TDM that most home and project studios don't need. Most of the people I've seen complaining about the price of HD are people that probably don't really need HD, but would like to know they have "The Flagship" in their home studio. And I noticed Peppertree never made any comments back about my asking if he uses HD a lot. I think that is the problem, people who don't use HD all the time, or even if they bought HD just because it was the "most expensive" product but don't need all that functionality, are the ones looking and saying "hey, that LE system can do everything I need and it's a lot cheaper...WTF!?!?!" For me, I don't own an HD system, but I work at full time at a studio that has 3 HD systems and 6 LE systems (and we also have Logic and Digital Performer) and I do freelance work as well at a lot of studios around the LA area. I can see the differences and the need for both... but I can also see where Native cannot inherently do things TDM can. And the things that allow HD to do those things are what make it very expensive.

Like I had mentioned, someone complaining that a TDM plugin is twice as expensive as RTAS isn't the same person that's debating whether they should purchase an M6000 or 960L, or possibly looking into a TDM alternative. Or are trying to decide if they should get a Manley Slam instead of a Vac Rac. Its a different type of user, different requirements, and different set of "ideals" that they feel are important.

Who knows, maybe the next incarnation of HD is native or at least has more native integration. But one of the inherent problems within a Native system (not just digi, but any native processing system) is latency. Sure you can get a box that does a hardwired mult to the outputs so you don't hear latency when tracking...or you can get an Apogee Symphony system for low latency tracking... but you'll notice there is absolutely NO solution that gives you low latency AND the use of plugins at the same time in a native environment. It's just not possible currently (that's not to say these obstacles won't be overcome in the future).

And that's a big deal...With everything else you are "compromising" to suit the flaw within your system. Those compromises might be ok if you have a home studio and you record your band, or a small studio recording local artists. But if you have clients that are traveling far and wide to be there, and are spending tons of money on the players, travel, hotels, etc, they aren't going to want to hear "no, sorry I can't do that...". They'd rather pay to go somewhere else. I know, I work with/for quite a few of these types.

And yes, they are the minority... but they are still a niche that digi has filled for years without any major competition. And in filling that niche, it has made them THE system to aspire to by all the novice users. To use the car analogy again, it just seems like a lot of people are complaining that they want the performance of a Ferrari but with the features of a Lexus. And then are complaining that they have to spend sooo much money to get a ferrari when it doesn't have all the amenities of their Lexus... and why oh why is Ferrarri punishing it's customers by not putting in fully automatic transmissions and fancy nav systems and cup holders, for god's sake how can anyone be expected to drive with only two cup holders in the car!!! And air conditioning, why do i have to pay EXTRA for A/C in a ferrari when my lexus comes with it standard!?! And why isn't the Ferrari a hybrid!? The lexus is hybrid, I'm sure they can make a V8 or V12 500hp engine a hybrid, I mean come on guys!?! Get with the program!!

Maybe you see it this way, maybe you don't. But when I look back over this thread and others with a similar gripe... that's what I see. I'm not trying to be mean here, or confrontational. Nor am I trying to belittle anyone.

[O.G. Killa]

Originally Posted by peppertree

All 100% absolutely true BUT you are missing the point which is HD vs. LE aka TDM vs. Native.

The OP is saying time to embrace native and I'm saying Digi already has a great native solution they are just crippling it to keep this HD exclusivity thing going. And they aren't going to undermine that exclusivity thing by pre-announcing something that makes their existing native solution look more like the future than their exclusive line.

I see what you are saying... I personally don't see it that way. It's more one of compatibility than anything else. Certain systems can't handle more than 32 or 48 tracks without getting errors. That's one of the ways they come up with these specs. On most of the older computers still in use, you cannot get more than 32~48 tracks at once without dropping samples. Some of the faster and newer computers can do more... But as far as I know, Protools is the only DAW that will absolutely not drop any samples and it will stop and pop up an error to tell you something is wrong before it does. I get session files from composers and producers all the time that work in other daws, and there are subtle clicks and pops in the audio files that the composers supposedly couldn't hear. Yet there they are, clear as day. And this has been happening for YEARS. It's not a recent thing. Before I go through and mix the project, part of my editing is cleaning up all these digital errors in the files... which are dropped samples. And I really feel one of the reasons digi is doing higher track counts is because you now have the option to turn off those errors/stops in the playback engine. So while PT can do higher track counts now, it will be subject to all the errors that happen within other systems and users just don't notice.

[peppertree]

I have worked full-time in HD studios and know my way about it well enough.

But I'm a different breed than is in your facile evenly split world. I have a pro studio that doesn't have HD and instead has high-end outboard for the same money. And I don't say "no I can't do that" because I just go ahead and do whatever "it" is with outboard...with lower latency and probably much more interesting, immediate results.

The intentional crippling of LE/MP is just nuisance to me. Galling and infuriating but just nuisance. There's nothing I can't do with my rig in practice...especially not with the Complete Extortion Toolkit I cursed as I paid for.

Digi could make my life easy if they wanted to in about a day. I've been asking them to make my life easy this entire decade and I'm not alone. I may have bought that toolkit for four times the price of Logic Studio and basically no additional features...but there's no way I'd buy an HD rig. The technology is dead and gone. Sure people are still using it, just like they are using tape! But unlike tape, there is nothing we'll ever miss from the HD rigs sonically.

And even the nuisances digi sticks in my eye all day long aren't as nuisance as operating a reel-to-reel...

[Bentley Ferrari]

[QUOTE=peppertree;1366927]

But I'm a different breed than is in your facile evenly split world.

peppertree...
postin' on the DUC.... reachin' out with understandin' an' acceptance... catchin' flies with... er.... keepin' it interesting...

O Provocative One:
I wonder if your world isn't a little evenly split; facile, even? Consider the following, if you will:

1.) Although you cast your argument as HD vs. Native, your argument is really Digidesign HD vs. Digidesign LE, which is much different. It is precisely the reliable functionality via the integration of Digidesign software and hardware (HD, LE MP) that attracts most of its users; at least that's what attracted me and keeps me.

2.) From point #1 we come to the next obvious question regarding HD's power and future development in light of ongoing changes in computer hardware and software. I don't see this as HD vs. Native, and my post beginning this thread was not cast as HD vs. Native.

Rather, I think you can look at it this way: Digidesign's "flagship" product has evolved as the personal computer and attendant technology has evolved. At some point, DSP cards became feasible because we had the right chips and computers came with the right slots, to put it simplistically. So HD wasn't developed entirely as an alternative to Native alone (although it certainly is that), but it was developed because advances in Native technology made it possible.

In the same way, I'm looking at the current and near-future advances in native hardware and software, and wondering how it will impact Digidesign's next steps. I don't see the logical necessity of your either/or scenario. Just as Digidesign developed DSP cards and interfaces that took advantage of technology in the 1990's and early 2000's, it ought to be able to develop a reliable, superior, integrated hardware/software solution based on the major advancements we're witnessing today, for example:

-- GPU based processing cards (these would be a quantum leap forward...gigaflops instead of gigahertz of processing power)

-- 64 bit apps (architecture, not sample bit depth)

-- a Venue-type stand alone system for the studio

I think the integrated approach is still key, no matter how much native power you have. Power without reliable, consistent, predictable application is of very little benefit to the kind of user that uses HD.

[peppertree]

That's an intelligent challenge, thank you Mr. Ferrari.

To tell you the truth I am not a DSP engineer and thus I don't know what digi's options are specifically in terms of DSP acceleration. If they want to use a bytecode system for their next-gen replacement for TDM plugins, that will just-in-time compile itself for a GPU-based solution they could do that, but it doesn't seem consistent with their go-it-alone, not-invented-here-syndrome corporate culture.

Their general strategy has been to standardize the platform on one or two basic hardware platforms and milk that for as long as they can. This has costs and benefits, the stability and excellent support is a benefit, the cost includes relatively rapid (and now accelerating) obsolescence.

As for 64bit my understanding is that it's just another tradeoff and many apps will run faster in 32bit. I don't know what parts of pro tools would benefit from what implementation. I have argued forcefully in favor of floating point for workflow and sonic reasons, and I figure they will move everything to floating point for the next generation.

That is also consistent with their challenge of being a platform vendor: right now, it's quite easy to port code between RTAS, VST and AU, but porting to TDM requires everything be rethought and translated into a different language. I don't think Digi is "must-have" enough any longer to get that level of effort from 3rd parties as it once was.

Now tying all those strands together with OpenCL (the bytecode system for running on GPU's) it certainly makes one think that Digi may in fact adopt much of that architecture purely for code compatibility reasons. If the whole audio DSP industry moves that way then Digi gets to sell its own custom hardware to run that bytecode but doesn't have the missionary challenge of getting 3rd parties to make special ports of their code to it. Right?

So yes I see that as possible, but the question remains, where is their sustainable competitive advantage? All the native DAWs (including their own LE) one must presume would run that same bytecode just on commodity GPUs. My understanding is that's an awful lot of crunch even in the absence of special accelerator cards. So will there be a practical need for people to buy even more DSP just for audio? Is enough enough?

Again a question is how low can realtime latency get without a special card that offers both i/o and DSP (like the current TDM cards do, and the Metric Halo 2D does). If Apple can work it so that we get realtime DSP under 2ms @ 44.1 then it's basically curtains for any special card Digi wants to sell us. People will just buy MADI cards (like the current RME ones) and hook their converters of choice (e.g. the SSL ones) to them for hundreds of channels of i/o as needed. But if those latencies can't be brought under 5ms or so, Digi still has some life left.

It would be galling if all the native plugins move to OpenCL and digi cripples that out of LE, like it has with today's DSP accelerators (UAD, LiquidMix, Duende, etc.). I have a feeling digi won't be able to do that anymore, because the 3rd parties will just drop RTAS support on them. They can't afford that; a DAW is really just a container of plugins with some editors thrown in.

Now as to a standalone system...Pro Tools is already closed enough! Some people like that standalone thing, like IZ RADAR, but the market has spoken on that and most people don't. The reason is because of the enormous value-add a general purpose operating system offers. Digi doesn't have to go do the missionary developer evangelism thing to get people to write to their closed platform....they just have to let people work with those apps, e.g. offering a ReWire plugin for Reason and Ableton integration into Pro Tools. I can drop a file I'm working on into izotope RX. Etc. A well-tended open system can retain stability!

I just read about someone dumping an Icon in favor of an analog console and it appears that trend is speeding up. Things are looking really bad for digi right now. They still have the standard session file format and they have the best audio editor. But the days where they could be dictators seem to be coming to an end. I will be stunned if they can pull something off that gives them dictatorial command of the market for another 5 years, with huge gross margins of yesteryear. I don't see an opportunity. I think they should get some humility, in a hurry, and "stand out by fitting in."

[Bentley Ferrari]

Maybe THIS (Click for article) will be Digi's next step.
The link goes to a story about Honda researchers
who've built a prototype machine that links
a machine's movements directly to the user's thoughts.

Kidding aside, we may live to see
the development of a system that
can read the sound in your head
and then record it.

No plugins necessary.
Mix in your head, too.
And never lose a song idea again.
Pro Tools 26.0, baby! Digidesign helmet required.