Community
Smoke Forum
Welcome to Autodesk’s Smoke Forums. Share your knowledge, ask questions, and explore popular Smoke topics.
cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Is Smoke 2015 really "optimized" for the new Mac Pro?

14 REPLIES 14
Reply
Message 1 of 15
ff
Explorer
1304 Views, 14 Replies

Is Smoke 2015 really "optimized" for the new Mac Pro?

I keep reading about all the bugs but I haven't seen anyone commenting on any performance/interactivity improvements. Are there any?
14 REPLIES 14
Message 2 of 15
email
in reply to: ff

I'm hoping the 'optimized' bit will be in Ext1. Does the release date have any connection with Yosemite?

 

Sven

 

Message 3 of 15
moreymichael
in reply to: email

I'm running 2015 on a brand new Mac Pro, fully loaded. Smoke ran better on my 2011 MBP. The Mac Pro has trouble playing back HD in real time; this has been investigated by several service calls by AD, with no answers. The interactivity is no better (and no worse) than either the 2011 MBP or the 15" 2013 Retina MBP that was had been using. I was expecting a significant, or at least noticeable, difference moving up to a computer with such power--alas...clearly, AD has issues to solve in talking to Mavericks, in order make use of all that supposed power.

 

If your current system is working, I cannot recommend spending the $$ to upgrade just yet. I'm waiting to see Nuke Studio. 

 

 

Cheers, Mike

Late 2013 MacPro, 10.9.5, 2.7 GHz 12-Core, 64GB ram, AMD D700, 1TB SSD
Message 4 of 15
ff
Explorer
in reply to: moreymichael

Thank you so much for the info. I suspected as much, and given the fact that Nuke Studio is weeks away I am really considering making the jump.

I waste an enormous amount of time creating proxies in Smoke (we deal with LOTS of mixed formats and quite a bit of 4k footage). The fact that in Premiere/After Effects/Nuke you can drop down the resolution on the fly to aid with the interactivity and you can't do this with Smoke makes no sense to me. Also, the software as it is seems very buggy and finicky. I guess the price really was too good to be true.
Message 5 of 15
mpixls
in reply to: moreymichael

"The Mac Pro has trouble playing back HD in real time; this has been investigated by several service calls by AD, with no answers."

Are you getting real-time playback from the HDMI output by chance?  Some users in here have reported that output of the HDMI is real time, and thunderbolt is not. I have only a thunderbolt display, and it does not show real-time, although de-bug shows 3x the frame rate I need.

 

"I was expecting a significant, or at least noticeable, difference moving up to a computer with such power--alas...clearly, AD has issues to solve in talking to Mavericks, in order make use of all that supposed power."

Agreed. I have the same system as you.  Interactivity is good, but the HWAA bug is still there, and as you commented I don't believe that we are getting anywhere near the performace out of the Mac Pro that other software packages ARE getting.

 

What concerns me is that I am not wild about the subscription model, and being a perpetual license owner a fix will come after the perpetual licence has expired.  A letdown to put it mildly.

Mac Pro 2013
2.7 GHz 12-Core Intel Xeon E5
Dual AMD FirePro D700
64 GB RAM
1 TB APPLE SSD
Promise Pegasus2 R4
Message 6 of 15
moreymichael
in reply to: mpixls

The HDMI doesn't seem to make a difference on this system, although the 15" Retina MBP would only play HD if I had the HDMI connected (nobody at AD could figure that one out either). And this problem on the Mac Pro is using a brand new Thunderbolt 2 SSD drive. Only happens on certain jobs, on certain clips. My current project seems to be working, but that's unusual.

 

I am also on a permanent license, which, for many reasons, I'm grateful for. I don't expect a resolution before my support runs out. But you never know. Fingers crossed.

 

Cheers, Mike

Late 2013 MacPro, 10.9.5, 2.7 GHz 12-Core, 64GB ram, AMD D700, 1TB SSD
Message 7 of 15
arichards
in reply to: ff

Sorry if this is slightly off-thread but here's my worry. What else does AD ser to "gain" from changing subscription to rental? When you subscribe to a magazine you get updates or you just keep the last issue. They update magazines monthly, etc. When you rent something you have to keep paying ot you get thrown out. On the former the pressure seems to be on them and on the latter on you. The pressure is rhus lessened on them to update as you bave to keep paying anyway to lose tbe last version. Now I placed scare quotes around "seems" as it really would be highly counterproductive (autoimmune, in my verbiage) for them to do this as people would soon loose interest, as some already are, i. Software that does not keep up. That The Foundry js going well ahead with NukeStudio and Blackmagic with Fusion really is great news in this regard. As I said in another thread and countless countless times, these other comanies are not peotecting any other products in their stable and so they're motoring ahead. smoke's latest updates seem to be really slowing down. If flame was cheaper and on mac I am sure that there would be much leveraging of the nMP (ahem background render and real dual support) but as it is you only seemed to get pathetically punished by AD for owning this stuff. This HWAA is really getting beyond a jone now ("its Apple's fault) and Smome on Mac just seems to be fading away. My eyes are focusing on other solutions.

Cheers
Tony
HP Z840, 80GB Ram, Quadro M6000x24GB
Message 8 of 15
moreymichael
in reply to: arichards

I'm with you, friend. The move from subscription to desktop is largely semantics, except for the fact it now costs more than twice as much, for exactly the same service, as far as I can tell. The "rental" thing works really well for Adobe, because for $50/mo, I will always have the most recent versions of PS, Illust, AE, etc., which was a problem before, when I got an .eps file from an art director who was using a newer version of Illustrator than me. So, good deal.

 

As for Flame on mac, that's what Flame Assist is. They're not going to give it to us. Period. From a business standpoint, they must know what they're doing. Making money. Long live Nuke.

 

Cheers, Mike

Late 2013 MacPro, 10.9.5, 2.7 GHz 12-Core, 64GB ram, AMD D700, 1TB SSD
Message 9 of 15
mikeparsons
in reply to: moreymichael

As an end user I can only speculate like every one else here. However I think I know what happened. 

 

Shared code.

 

In the early days of smoke on mac AD would have looked at OSX, seen how close OSX was to Linux and worked out a way to port their code to Mac without hideous amounts of fine tuning. They decided this was cool and saw an opportunity to increase sales to a less aff;uents section of the post community.

 

Then Apple screwed them. No new mac pro for years. Smoke 2011 got a few new features but the users were getting restless, new systems like Nuke offered so much power for half the price point. Then Apple introcued FCP-X to universal hatred. Everyone started to leave the sinking ship that Apple pro video was seen to be, no new macs and now iMovie pro? Time to go. AD already must have been working on CFX and the new batch engine but along with that jumped into a GUI overhall to make it more FCP like. They also went for an unprecedented level of end user interaction with a public beta. AND they addressed the price point issue,making it LESS than Nuke and Half the price of Nuke X literally reversing the price differential. Immediately Smoke on Mac buzzed, for the first time in a decade people who were not already customers were excited about AD finishing products.

 

I too was more excited than I had been for a long time, my new group of friends and colleagues on the Area forum came from a much more diverse cross-section of the industry and so their experience, needs and ideas were literally a breath of fresh air. But then reality hit. Turns out it takes a long time to completely redesign a GUI. The beta was very unpolished in all the new areas that excited the new users. Editorial was weak, no trim, no audio scrub, no multicam, hmm surely this couldnt be the killer app AD promised. 

 

Additionally these new users whilst not as old grizzly and pushy as the old ones weren't as tech savvy. Mac users dont peak under the hood like Linux/Irix geeks did. Backburner was the silver bullet that derailed Smoke on Mac 2013. Shared code turns out to have been a poor solution for mobile systems with changing IP addresses... 

 

Then Apple screwed AD some more. Lion changed the way Macs identified their mac address seemingly randomly assigning ethernet identifiers on boot. This screwed Pro Tools, it screwed Maya and it screwed Smoke. Rebooting would occasionally 'fix' it only for it to break again on the next reboot. Aaaaarrrrrrgggghhhhhh.

 

I took the time to write a simple guide to avoiding the backburner issue but for many it was once bitten twice shy - understandable - we are an industry where delivery is all that counts and people were missing deadlines and losing clients.

 

Then Adobe jumped into the fray. Premiere was picking up market share as people moved away from FCP and avoided FCP-X. Seeing an oppotunity Adobe introduced our industry to leased software. The market was stunned. Which is odd as when I started in 1984 a lot of software was leased especially on mainframes but I guess memories are short... Anyway the result of the Adobe deal was simple. They leveraged the ubiquity of Photoshop to sell (lease) everything else. No matter who you are in video you have to have photoshop. Soooo if you get an editor and effects for free may as well take a look right?

 

Premiere had been improving for a long time but most people had no need to look again, happy as they were with their FCP or Avid. But excited and happy with their new creative cloud (nice marketing given that you download everything and run it locally - kind of the OPPOSITE of cloud computing but whatever) the newly rose tinted glasses fully on people liked Premiere and hey it links to After Effects and I can even one click limit my audio to ITU 1773/3 and create broadcast deliverables such as MXF-op1a (not to mention cinema DCPs)! Adobe jumped from an also ran in pro editing to the first choice for a new system or a new editor. Its wide codec support and the fact that Adobe has 25 years more experience creating cross platform software was a solid bonus.

 

Meanwhile in camp AD their core clients were angry. Old people like me were harsh critics of the new GUI and the compromises made at the new price point... A wacom at 500 bucks is a significant cost in a 3500 system yet a post it note on the end of a flame purchase. Attempts to woo the FCP refugees led to a new set of keyboard commands and lack of love for the loyal old users and so began the kicking of AD. I as much as anyone tried to snap them out of the dead end they were running towards. Sacrificing speed to gain new users was terrible enough but i also started to see frame drops and quality issues in default and CRASHES. Losing a days work, inability to export, caching issues and more stopped all love for the company - the development team were working hard but there was just so much to do and emotions ran high as we waited for the version that fixed everything. 

 

The Apple screwed them. Mavericks came out. New methods of writing to the screen. The first steps away from the quicktime engine. Background transcodes of all video clips to H264 and a less scriptable core video engine. Enough to stop Smoke working. Enough to stop a lot of things working. Adobe just tweaked and moved on. But AD was screwed because... shared code. 

 

Ok the mavericks fix will take some time, and right now AD is 3 years into their new Smoke, time to show the old users some love as the new users they hoped for just arent arriving in significan numbers. But the new GUI has been committed to and drops onto Flame users heads. Bam. Angry voices. From people who have been their most loyal customers. Why? Because flames cost a ton of money and are the backbone of the post house establishment. Learn a new GUI to work slower than I can already? Ain't nobody got time for that.

 

The hard engineering in new flame was great. The reactor engine fixed many render inaccuracies. The move to half float was long ovedue, long standing issues such as corner pinning errors in 3d space were also fixed. But the ungrateful sods that we flame artists are meant AD just got kicked more. Yes - the new quality is nice but why do I have to press a button on the screen to gand a clip when I used to just press g? I can only imagine the hatred in the AD dev room.

 

Brian was a constant reality check of 'guys this stuff is actually improving' but many smaller owner operators like me and most of the post houses simply didnt upgrade. The mantra of its just a few more clicks enraged me particularly as I know that those few clicks add up and even losign just an hour a day is long enough to make the difference between seeing the kids when you get home and them being in bed. It can also means the difference between being married and being single.

 

But along the way AD kept at it. Changes in management founding VP Marc Petit leaving, massively experienced people sent in to fix it. A doubling down on software development and a market analysis. And then Apple screwed AD.

 

Heres the new mac pro. None of your Smoke users can keep any of their hardware. Throw the raid in the bin, throw away your quadros, throw away you kona cards and start again. Oh and by the way in total reversal of the last 5 years we are using ATI cards and open CL. Boom.

 

Just as Avid had to drop macs when the power pc came out with too few slots for the ABVB /3d genie/ raid card and targa 3000 so AD had to reevaluate. While we all complained. A lot.

 

The main thing about Open CL is it isnt Open GL. Its also poorly implemented under red hat. So that meant the end of shared code. Which is both a good thing and a bad thing. If I had been AD I would have thrown in the towel on OSX. Sod you all, all you do is moan anyway 🙂

 

But they didnt. What they did was to study what the needed to do to compete in a very different landscape to 3 years earlier.

- Open CL, major rewrite

- quicktime imminently being abandoned, cant depend on apple for codec support like in the past

- adobe subscription model is a huge success

- needs even lower price point based on FCP-X gaining traction, Lighworks, Premiere and Davinci resolve

 

And thats where we are. They kept going. A subsciption model (which they cant even call that as maintenance is already called subscription... I know lets call it desktop) to compete financially, a seperate code base optimised to the new boxes and OS. We cant expect them to write flame twice. So the new desktop version has to make some comprimises. Some wont be noticed some will. And the reintroduction of sparks shows they are still listening to the complaints. Until Apple screws them again.

 

Watching all this from the sidelines is the main reason I stopped complaining. AD has had a terrible time with a large part of their perceived failings being totally outside their control. They have reached out to me and many others with senior management and proiduct managers makign skype calls, flying to bangkok just to meet a few flame guys, flying to LA, NY and London to meet people there. Their commitment to this industry is solid. I have direct regular contact with the highest level now in a way I didn't enjoy even in the Discreet glory days. And still the dev team press on.

 

The new version of flame featured reactor 2 with background rendering on a second K5000 card. Multiple streams of realtime 4k on the z820 etc. They know what they are doing in the hard engineering stuff - its just been impossible for them to show it on the mac as Apple keep pulling the rug from under their feet.

 

I dont know if I think Smoke desktop is going to work for them but I do think its going to outperform Smoke 2015 going into the future. Flame compatability is an issue for some but thats more of a general complaint limited really only to old freelance users who prep at home then dry hire a room to finish. In reality if you have a flame you need to be compatable with you can have flame assist, if you dont have a flame then you are better off with smoke desktop. Dropping sparks on the desktop was I believe the best decision but they needed to have had OFX support on the same day. So sparks are back. 

 

What needs to happen now is four things.

 

We need to trust them in the flame world, and reactor 2 went a long way to regaining my trust. All I've ever wanted is raw speed. I am consistent that way 🙂 The latest flames are the fastest I've ever seen.

Now dropping shared code means we can get the GUI back on track. We know every flame artost has a tablet so lets refocuss back on gestural interaction and speed.

 

Secondly AD needs to find a way back in with Smoke desktop, a way to reignite its use. A way to get new users excited again. I have some ideas, the easiest of course is to just give 6 months free licenses out with every 3d package purchase and to every registered 3dmax, Maya and XSI user on earth. That should convert many users in 6 months time. Then they need to get everyone here on board. Maybe a similar 6 month deal as a cross grade from your perpetual licenses because trust me the desktop version is going to be better than hanging onto you perpetual license going forward. Lastly they need a crossgrade program for other users. Maybe half price subscriptions if you come from another editing system. The foundry summer sale just did this - once I sent proof of ownership of after effects I got 20% off some nuke seats...

 

Thirdly, freelance connectivity needs to be addressed. Either let flame owners 'sponsor' a user with a temp license for flame assist or simply add a 'flame export' module to smoke desktop. Maybe a smoke desktop import module for flame assist... I dont have the answer but there ahas to be a way to do stuff at home in a cheap system that you can use in the big rooms later.

 

And fourth, they need to extend their marketing beyond hype into a more grassroots hands on approach. im thinking FCP user groups style. Roadshows in towns and villages, converting the masses by word of mouth and example. This has to be done socially via groups like SMPTE, the DLF, and ACE. Give Brian a job and pay him for his enthusiam, form advisory panels in each major market and incentivise them to spread the word. The biggest hurdle AD has is that there are few flame artists under 40. We are old and about to start retiring/dying and we have to be replaced. Nuke studio has mindshare now, smoke needs to reclaim it.

 

So there you go - my 2 cents worth. I've been angrier with AD than I ever was with Quantel (or even CMX when they released the CMX 6000) but I dont think its because they are the evil empire or 1%ers getting rich while we all suffer. Things just happen and we all go to work for the same reason wherever we fit in the grand scheme. They are trying hard. They are doing the best they can and Im seeing improvements so I'm trying hard not to complain so much. Now if they can make the roto tools as good as the ones in Nuke I may yet start saying nice things.

 

best regards

 

Mike

All's well that ends. That's why its called finishing.
Message 10 of 15
moreymichael
in reply to: mikeparsons

Wow. Mike, thanks for the insight. Glad to know there are a few of us on the "inside". Do they meet you in parking garages and refer to you as "the asset"? Smiley Happy

 

I do like the subscription model. As I mentioned above, I subscribe to Adobe, and if I were more ambitious, I'd probably figure out how to use AE, and then I'd have a complete compositor's toolset for $50/mo. You've convinced me that AD's move to "desktop" is more than simply a money grab (although I still choke on the more-than-double annual price point). But, regardless of whether or not I continue to use AD products, I will almost certainly  join the ranks of NUKEists because it's clear to the end user that The Foundry cares deeply for the compositor. Maybe that will change in the future, and then again, maybe AD will unleash a more "discreet" attitude towards its core user base. Who knows.

 

One way or another, it would be nice to know that I haven't shelled out 13K for nothing (I'm freelance, so that represents a lot of mortgage payments). Even if it's Apple's dirty trick, AD is certifing their product on that gear, and leaving us all scratching our heads after the fact when it doesn't work. 

 

Cheers, Mike

Late 2013 MacPro, 10.9.5, 2.7 GHz 12-Core, 64GB ram, AMD D700, 1TB SSD
Message 11 of 15

Mike, as always a great post, very detailed and still spot-on.

 

But with Flame Assist and now Flare on the Mac the shared code-issue won't dissappear. Instead, it will be there with every release. Now that the Mac is fully integrated in the Flame Family, we'll be more depentend on proper code on both platforms. Something like Smoke for Mac 2014 should never happen again. I don't necisseraly need them to support the latest and greatest OS X from day one. If it takes until 2015 to support Yosemite, then so be it. I can live with that for sure.

 

Cheers


Christoph

 

Message 12 of 15
julioleon
in reply to: ff

Thanks Mike!
Smoke 2015 SP3
Mac Pro 6 cores
OS X 10.10
32 GB RAM
Message 13 of 15
ksirul
in reply to: ff

This combo has caused my nothing but grief.  I have found that it is workable on short form (5 minute pieces or less) but I am mostly working on films that are 40 minutes to over 2 hours.  The system is extremely slow and unworkable.  Especially when text on gaps are involved.  Once my  current 2 films that are on it are done, which could be another month unfortunately due to additional network formatting and DCP prep, I'm done with it until these slowness issues are fixed.  One major issue that AD support has found is that Mavericks does not allow Smoke to use more than 4gb or so of RAM which is much of the problem.  Many other issues, too long to list here (I've mentioned many in past posts).  I've been told by a couple of AD folks that these Mavericks issues are being worked on but no one knows exactly when they'll be sorted out.

 

KEN

Message 14 of 15
ff
Explorer
in reply to: ksirul

Thanks to everyone for the replies. The company I work for is about to make some upgrades and the info I?ve gotten from you guys is really helpful in helping to inform our decision. I can see how many of the problems Smoke is having can be Apple?s fault, but on the other hand, MANY other pieces of software work perfectly well on Mac and at least COME CLOSE to fulfilling their marketing claims. It?s a shame. I?ve been a Flame/Smoke user since 2005 and I can honestly say that unless you?re willing to fork over the price for a full Flame system, I don?t think Smoke is worth even the price they?re asking for nowadays. I?d rather spend $9k on a Nuke system that works (and yes, I?ll work around the shortcomings of that software with the addition of Premiere) than to pay $1,750/year on a piece of software that makes me wait HOURS while creating the proxies necessary to work with large formats while it does this in the ?background? (yes, that is laughable) and have the addition of MANY other wonderful bugs. Not to mention the misrepresentation of being optimized for a very expensive piece of computer equipment which some people (me included) was taking at face value. Goodbye Autodesk, it was great while it lasted.
Message 15 of 15
ksirul
in reply to: ff

I do have to say that the AD tech support team has been great, but these nMP/Mavericks issues are beyond their immediate control.  I too was burned by buying the nMP based on being told that it would work with the 2015 version.  For that I blame myself for buying into sales hype.  I've been doing this long enough that I should have been smarter about that, but I needed a replacement for my DS pretty quickly, which, until Smoke starts working on my new MP or I can recoup the money I spent on the Mac to buy a Z820 for a different system (Nucoda maybe?), I'm back on DS for a bit.

Can't find what you're looking for? Ask the community or share your knowledge.

Post to forums  

Autodesk Design & Make Report