Adobe Responds to Intel-Only Soundbooth

[phasornc]

I heard Soundbooth is based on a Windows-Program Adobe aquired, a peace of software with heavily optimized x86-code. If this is true, it would need huge efforts to create a universal binary, not just a click in Xcodes compiler options.

Before people freak out, audio is a very different beast from every type of application out there. Audio apps need to do a tremendous amount of realtime processing, effects, synths, multi-track playback and at the same time record multiple live tracks and record live controller data. And all that stuff has to stay in sync. Many audio apps use parts that are programmed in assembly or at least programmed in C using hooks into CPU specific calls like Alti-Vec or SSE3. Apple tried to help out with the Accelerate framework to give developers one grammar that can translater into Altivec for PPC and SSE3 for Intel. I'm not an expert but I bet the framework is fine for playing back a stereo track or 6 track movie with Video, but how well well does it hold up with 24 tracks loaded with real-time effects? Probably not that well. Just like Apples VST to AU conversion SDK which helped out with some simple plugins but couldn't come near the depth needed for automatically converting some of the more complex synthesizers.

The point I'm getting at is that Soundbooth was developed by the former Syntrillium CoolEditPro team that Adobe acquired and changed it's name to Audition. These peope were an X86 team only, writing X86 exclusive code. Should they devlve into the "Accelerate Framework" only to possible run into a brick wall 6 months into development and be forced to rewrite code SSE native? If they use they Accelerate Framework they need to code bases, one for Windows and one for Macs. By saying "Intel Only" they probably just recycled the same X86 C or C++ code that they used in Windows and potentially someone or some group from the Macromedia team that used to make SoundEdit 16 or Deck created a Carbon timeline gui for the new C processing code.

So before you diss Adobe, thank them for giving the SoundEdit "guys" new purpose in their lives, because the alternative was no Mac version at all. Like I said Audio is a very different beast and I suspect this will be the first of many Intel only Mac audio apps. Indeed now is a great time to be a Mac gui coder.

Michael

 

[Fireburst]

I agree with Adobe. Pointless investing a new product to a dead architecture. The only important factor is PPC support for CS3 for those still awaiting the chance to upgrade.

 

[scottlinux]

^ Ever heard of this:

http://labs.adobe.com/technologies/lightroom/

 

[Sun Baked]

It is bound to happen, as application migrate forward.

But this is rather quick -- and not a bad thing.

If it is a must have app, people will upgrade.

There is a cost to having a 10.3+ app versus a Intel/10.5 only app -- and that is being able to use all the current features available.

 

[gnasher729]

True indeed!

It makes sense for new applications of this magnitude to be Intel only, I suppose. But what tools are they using to build this? If they can make Universal Binaries of their other applications, isn't it simply a recompile for PPC? Or is all about the optimizations for the hardware?

Actually, it doesn't make sense. The only problem changing PowerPC applications to Universal applications was that the programmers believed that there code would forever run on PowerPC, so nonportable code was written. However, if you know ahead that your code should run on PowerPC _and_ Intel, only a completely incompetent programmer would write code that runs on one processor only.

And for a brand new application, you _don't_ optimise for the hardware. Time to market is the most important thing, optimisations can come later.

 

[iMeowbot]

^ Ever heard of this:

http://labs.adobe.com/technologies/lightroom/

That's a whole different animal. That program began life PowerPC (and Mac) only project, Universal and Windows support were added later.

 

[erikistired]

it's hard for me to call ppc dead when apple is still selling them (altho down to refurbs now it seems). however, if adobe wants to ignore a large installed user base, that would probably be their loss.

 

[iMeowbot]

Interestingly enough Soundbooth has some universal frameworks in it in according to a post here:

http://episteme.arstechnica.com/eve/forums/a/tpc/f/8300945231/m/461009861831

So somewhat of a sloppy job, that could easily be universal.

Most of that universal stuff looks like libraries borrowed from other Adobe products.

The real problem is the SSE instructions. No, the Accelerate framework is not some kind of cure-all here, because that framework only exists on OS X. Even if Adobe used Accelerate, it would be a port of the critical code to a new architecture (OS-specific rather than hardware-specific, but it's still a porting issue). SSE code, on the other hand, will work with both Wintel and Macintel.

 

[lkrupp]

No more innovation...

And so the PPC macs being obsolete begins. Glad I ditched mine and went intel.

The thing that concerns me most is the fact that the entire personal computer market is now locked into the x86 hardware architecture. We can now forget about any hardware innovation for the future. Intel now controls the future of hardware and, unless they approve, nothing will ever change (AMD is just the tail of the dog). The Wintel duopoly today is like the AT&T monopoly of yesteryear. Innovation, if you want to call it that, will come at Microsoft's and Intel's discretion when and if they the choose to do so. And there's nobody out there to say otherwise.

I said it before and I'll say it again. Adobe will eventually drop all OS X development and go totally Windows. Apple will have no choice but to either go it alone with productivity applications or switch over to Windows and become a Wintel OEM. Boot Camp, Parallels, Crossover will serve only to hasten the inevitable. Apple will have to do it or die.

 

[grahamtearne]

Im sorry but I have to disagree with several of your points. Firstly AMD was always more of a threat to Intel than the PPC market. Macs were the only computers to use them, marketshare was so low they wasnt a threat to intel at all, the only way ppc would have became a threat to intel is if there was a ppc version of windows.

What brings you to think adobe will drop os x? If they were why drop just ppc support and make x86 apps? they wouldnt bother dropping just 1 platform (ppc) and continue another (x86) if there long term plans were to drop support, its a waste of time and money supporting x86 for them to drop it altogether. By them supporting x86 shows they are looking forward.

The thing that concerns me most is the fact that the entire personal computer market is now locked into the x86 hardware architecture. We can now forget about any hardware innovation for the future. Intel now controls the future of hardware and, unless they approve, nothing will ever change (AMD is just the tail of the dog). The Wintel duopoly today is like the AT&T monopoly of yesteryear. Innovation, if you want to call it that, will come at Microsoft's and Intel's discretion when and if they the choose to do so. And there's nobody out there to say otherwise.

I said it before and I'll say it again. Adobe will eventually drop all OS X development and go totally Windows. Apple will have no choice but to either go it alone with productivity applications or switch over to Windows and become a Wintel OEM. Boot Camp, Parallels, Crossover will serve only to hasten the inevitable. Apple will have to do it or die.

 

[Counter]

NOW is the absolute best time to sell your PPC Mac.

 

[bousozoku]

So explain Lightroom being Universal?

Thank God it is Universal, even in it's beta version it's a fantastic app, and incredibly stable (far more stable than many final version apps, including Excel).

Lightroom was a Macromedia project, not an Adobe project.

The whole thing about Southbooth being easier to develop for a certain computer architecture, especially from scratch, is a joke. They're making the investment to create the application for Mac OS X, adding a separate PowerPC target isn't much work at all.

It comes down to this: they don't want to work at making it run well and know the the faster Intel processors will take up the slop.

 

[gugy]

NOW is the absolute best time to sell your PPC Mac.

It depends, If you relly on Adobe CS and After Effects I would wait until the next revision of the Mac Pro. The current Mac Pro is not as fast as the PM Quad G5 for CS2 and AE. Close, but not as fast.
Next revision of the Mac Pro will be probably a little faster under Rosetta. Then, I think is the right time to move since is said by Adobe that Intel native CS will be here next spring. then the Mac Pro will kick a**!

 

[Sun Baked]

It comes down to this: they don't want to work at making it run well and know the the faster Intel processors will take up the slop.

That is usually the case, look how much Office slowed down when they ditched the Mac-only code.

They did it to reduce support costs, and many of the bugs that came with the multiplatform version have yet to be fixed all these years later.

Sometimes it isn't a decision made by the programmers, but something stuffed down their throats by the bean counters.

Could simply have been a simple, hey we can save $50-250k by ditching PPC support.

 

[bousozoku]

That is usually the case, look how much Office slowed down when they ditched the Mac-only code.

They did it to reduce support costs, and many of the bugs that came with the multiplatform version have yet to be fixed all these years later.

Sometimes it isn't a decision made by the programmers, but something stuffed down their throats by the bean counters.

Could simply have been a simple, hey we can save $50-250k by ditching PPC support.

Sure, once John Warnock stopped leading the company and Bruce Chizen took over, this kind of thinking came front-and-center. One had a passion for creating computer software and the other has a passion for money.

They want money and they're not taking "no" for an answer. It's obvious with Photoshop 7 and Illustrator 10 that they did the most minimal job that they could do to make them run in Mac OS X.

Many companies might consider re-working the code causing most of the trouble, in order to save time later but Adobe went ahead and added features instead of fixing problems because they could charge for one and not the other.

 

[SMM]

So explain Lightroom being Universal?

Thank God it is Universal, even in it's beta version it's a fantastic app, and incredibly stable (far more stable than many final version apps, including Excel).

Excel - stable, not two words I usually see in the same sentence.

 

[SMM]

NOW is the absolute best time to sell your PPC Mac.

Depends on who you are, mate. As long as Apple keeps supporting PPC with FinalCut Suite (or derivative), I will keep using my Powermacs. I would love to see some panic and fire sale on G5 Quads.

 

[bzgnyc]

An Adobe blog by John Nack addresses some recent concerns due to a recent release of the beta of their new SoundBooth application which is available for Intel Macs only.
:

I agree with their decision except for the problem that their big apps (e.g. CS2) are PPC-only and I think high-end users are waiting for CS3 before getting a Mac Pro. For high-end users, CS2 running on Mac Pro via Rosetta is not good enough. While I would guess that the number of end-users who use the CS2 for graphic design and SoundBooth for audio is small, it seems to me that SoundBooth would fit right into the Adobe Production Studio for digital video & audio. A high-end digital video/audio user would have to have an PowerMac G5 for everything but SoundBooth and a Mac Pro just for SoundBooth. While the Mac Pro would eventually become the primary computer once the main apps go Universal, everyone knows dual desktops is no fun after a few weeks not to mention a few months.

Adobe is jumping ahead on the Intel-migration. Until Adobe can offer Universal binaries of all their apps, its too soon to offer Intel-only versions of some. Its understandable that Adobe doesn't want to incur the extra cost of a Universal binary especially for a program such as SoundBooth which may have many process-specific optimizations. It is leaving their high-end audio/video customers in a pinch.

P.S.I am not a big fan of the IA32, but as IBM said the ISA is dead. Its not that innovation is not going to happen in processors, its just not going to happen at the ISA level. At least not for desktop and servers (its still possible for phones, PDAs, supercomputers, and not invented yet devices to go in a different direction).

 

[monke]

Will CS3 end up being the same? If so that would be insanely annoying. I'll find a meteor, tap into in, and remote control it into the Adobe headquartes. After Microsoft is destroyed though.

 

[EagerDragon]

I guess they have no interest/plans of selling the software to the millions of current PPC Mac owners.

 

[gugy]

Will CS3 end up being the same? If so that would be insanely annoying. I'll find a meteor, tap into in, and remote control it into the Adobe headquartes. After Microsoft is destroyed though.

I doubt it. They will have new features, plus I heard the whole program will be redesign from the ground up. That means a much faster and clean version.

I can't wait to see the benchmarks of CS3 with an octo-core mac pro!

 

[iMeowbot]

Just to set the record straight, Lightroom didn't come from Macromedia. It was very much an Adobe project.

 

[stephenli]

"For intel macs only" is far better than "For Windows only". Right?
remembered Adobe Premiere Pro?

 

[Mavimao]

I think it sucks for people who spent a lot of dough on a last gen g5, but for people like me and my setup (see below) a computer upgrade is definitely foreseeable and I, personally, have no real problem with a computer company making an x86 only app. I probably wouldn't be able to run it even if it was a universal app!

 

[h'biki]

I guess they have no interest/plans of selling the software to the millions of current PPC Mac owners.

... and how many PPC Owners *are* interested in a product like Soundbooth?

Their target market is going to be bulk buyers, like Radio Stations, Universities, Audio Post etc. They have high technology turn around and will likely be INtel by 2008 - which is like into version 1.0 of the company.