Smoke 2015 and Flame Assist - What does it really mean - Blog Post

[BKM]

Thought this might shine some light on things.

http://www.premiumbeat.com/blog/autodesk-smoke-2015-and-flame-assist-on-mac-what-does-it-really-mean...

[Anonymous]

Thanks for this, the write up does shed some light. However the video is private 😞

[Anonymous]

Thanks Brian.

Autodesk, will there be a cross grade path to Flame Assist for non Flame customers who are current Smoke subscribers, or does this imply that after Smoke 2015, if I upgrade my Smoke to (say) 2016 or whatever comes next, I'll not only lose Flame compatibility, but also the ability to use Sparks (about $4K worth in my case)? I work in an environment that has Flame, but I don't own it.

[arichards]

I think the idea is to cut off communications between Smoke and Flame by offering Flame Assist (first one free, next one more expensive than Smoke) as the bridge.

The cutting off of sparks to me is going in entirely the wrong direction as they should be opening up Smoke even more rather than shutting it down even more. Smoke is turning into an island when it should be trying to create an ecosystem. yes there are matchbox shaders but are they replacements for all those OFX plugins that every other software vendor seems to be embracing? When other software seems to be opening up more and more Smoke seems to be closing up even more than it was. For those who live in the UK, it seems that Smoke is following the Marks and Spencers model when it really should be seeing itself as a Waitrose! Then there's the introduction of background rendering on Flame via a second gpu. This is something that I've wanted for ages on Smoke and now I cannot ever see it happening. They're removing wire remote rendering from Smoke to protect Flame even more, so I really cannot see them introducing background rendering to Smoke anytime soon. Looks like my second D700 graphics card is never going to see the light of day in Smoke, whilst in other competing software it will. I wonder what The Foundry's announcements are going to be?

I must admit, I am beginning to lose hope...

Cheers

Tony

[Anonymous]

Brian,

Thanks for the clarifications.  I have some of the same questions everyone else seems to have...

1) Will there be a route for Smoke on Mac users to move to Flame Assist?
2) How will the feature sets change over time in your estimation? Flame Assist will be more comp/finish centric?

3) Will Smoke evolve into a more fcp-like application or stay in the "finishing" realm?

Again, thanks for what you do?  Don't stop Smoke tutorials now that you're getting Flame.

Artifact

[BKM]

---

@Anonymous wrote:

Brian,

Thanks for the clarifications.  I have some of the same questions everyone else seems to have...

1) Will there be a route for Smoke on Mac users to move to Flame Assist?
2) How will the feature sets change over time in your estimation? Flame Assist will be more comp/finish centric?

3) Will Smoke evolve into a more fcp-like application or stay in the "finishing" realm?

Again, thanks for what you do?  Don't stop Smoke tutorials now that you're getting Flame.

Artifact

---

1. Currently Flame Assit is only availble if you aslo have a Flame License.

2. Not sure. But one could guess that Flame Assist will be there to support Flame and with the Linux UI, I would think more interop workflows would be shared. Flare is still more comp centered as it has full batch. Flame Assist would conform, prep, and version. A well as I/O and Archive. Things that would free up Flame if they were done on Flame Assist.
3. Autodesk seems to want Smoke on the Mac eco system... that is.. you an use it as a finisher by importing XML, and now export an XML that points to media to say, use in Resolve.   The Effects will still be there. Not sure how it will spererate itself. 

And as far as not having Sparks. Sparks are dead and old. Matchbox is much better.  And now with the encrypted shaders (.mx) you may find some plugins for sale by 3rd parties. We can hope.  

[Anonymous]

Sparks cost almost as much as the host so the takeup for sparks is probably really low.

Matchbox is awesome.

GLSL code is all over the interwebs - you can see some of it in some of the matchbox shaders.

Collaborative matchbox shader writing would be a good class - maybe you could get the gov to float you some dollar for their year of code push.

Smoke on subscription is what people on smoke have been asking for.

(Certainly cheaper than the 15K I paid for it, & all my subs.)

I can't remember many flame people asking for flame on subscription - one or two maybe but not many.

I can't remember many flame people asking for flame on mac - one or two maybe but not many.

BG reactor - the 2nd GPU thingy - is not like multi GPU setups with other software - think of it more like burn in your workstation.

Burn is free if you're on subs - you just need to buy the burn node, network it blah blah.

BG reactor is way more expensive than a standard burn node BUT it's in the same box - no networking delay - no separate box.

You have to weigh up wether or not that is better for you than the cheaper option.

So would someone want this for smoke?

Yes if money is no option.

But Smoke on Mac automatically dictates that money is an option - if it were not you'd have flame premium on your desk.

I don't know what your setup is - college, institution, private tutor but why not try and work with autodesk and get yourself a flame premium, some flares and some flame assists and teach people how to work collaboratively?

Those helpful happy blokes and girls in London are good at problem solving - give them a call

[arichards]

Please don't take too much offence at this Brian, but this is the side of you that I just can't take seriously. I am a huge fan of the work that you do to help users of Smoke to find their feet, but as an advocate of the planning and as a mouthpiece for the organisation, I really have to take a breath. I am not so much bigging up Sparks but decrying the lack of proactivity on the part of Autodesk in opening up the software. I was very very clear in my post that this is what I meant. I am not a fan of Sparks but I am very envious of the shelves of produce that other products offer their users in terms of stocking produce from other manufacturers or plugin vendors. I don't love Waitrose so much for Waitrose's own brands but also as an infrastructure for the stocking of produce from other manufacturers. I am head over heels in love with the way that Smoke handles things, but also head over heals in hate for the way that it is so buttoned up and protected. The two sides of my love and hate are at war, but closed view that Smoke takes does not leave me happy and does not fill me with very much hope for the future. 

The worst news really was hearing about the one machine background rendering that is now available in Flame. That is 'exactly' what I want in Smoke. I really don't think that this is ever going to happen. I see SMoke becoming even more closed in than it ever was. The removal of Sparks is simply one symptom of a much larger disease. It is not itself the thing, it is the thing that it is a sign of. I hope this clears up this misinterpretation of my post?

Cheers

Tony

[Anonymous]

Oh & the other thing about BG reactor - you need two NVIDIA cards - only the Q6000 and the Qk6000 are supported so far - on current workstation hardware.

Since the plan is to support the latest, greatest, fastest, mostest you have to wonder how that fits into the Apple,ATI model?

[arichards]

But I am not complaining about what is available but about what is not available, and I am speaking as a beta tester and as somebody who has put in requests for items over a number of iterations of the software. I know the model that exists very well but I also know the model that I would like to exist. I see patterns forming over a number of iterations and I am interpolating a certain trajectory that is not, for me, the ideal place. I know very well the model that the linux boxes are following and of their difference, as stands, from the mac based ones.

Cheers
Tony

[Anonymous]

What did someone write last night?
The first rule of fight club is...

[arichards]

I think the nda covers discussion of what is within not discussion that you are in! 

[BKM]

I actually think this may help openthings up for Smoke. Since it won't be tied to the Flame ecosystem, it can live on it's own.  And not be in Flame's shadow. It will still be powerful, still have ConnectFX, and now is cheeper and more accessible to more people.. People who don;t need or care about Flame.

Since Smoke is a rental now, they need to keep up the features and workflows so that people renew each month.  

The pressure is now on Autodesk to do that and I think they know that. 

[arichards]

Yup the pressure is on and there is so much movement in the software world. I am not sure where I and my institution will be in a few years, but I am not liking the indications so far. I hope that you are right Brian, I truly do, but this just does not make me feel it. Time indeed, as it always does, will tell and reveal all. The trouble is, once it does there is no going back.

Right now I am beginning to look into other solutions, just in case I really do not like where it turns, as other software begins to turn where I like!...

Cheers

Tony

[Anonymous]

So a few people have asked what I think, seemingly expecting me to rant!

On the contrary I think this is a very shrewd move by AD. Ive been saying since the beta that it was all getting too cheap to be sustainable and I see this as a consolidation back towards systems. This means Flame assist can be as complex to setup and maintain as its always been with networked backburner remote rendering and so on secure in the knowledge that it will be setup by an AD systems guy.

A comparable place to look is Avid, they too are refocussing on bigger things, ISIS and cloud based offerings in the tens of thousands rather than playing in the FCPX 300 dollar space.

I am actually in the process of purchasing a Flame with several flame assists, right now as the new version on the new z820 is going to be amazingly fast and being able to properly integrate with assist stations excites me and makes the possibility of my all AD facility a bit closer to reality. I personally would like to see a Linux version of flame assist with flare rolled into it but I'm sure that will come - there really isn't a good reason to keep developing flare at this point. 

Its no inconsequential thing that the new flame assist version of smoke will be in traditional livery. Don't underestimate the degree of speed that familiarity of color coding will add. Now we just have to hope it gets desktop action and I will be very happy to finally retire my Smoke 2012s.

The rest I'll review when i get my hands on it, its easy to make a demo look good, but its mostly back to what we've all said on this very forum - it had better be stable from day 1 or no one will care. 

The only downside I see is for existing smoke owners who dropped off subscription. But then they have the option of cheap subscription model - the only real unknown is how limited the rental version will be going forward.

Also, the biggest expense for post houses will be reprinting their Smoke artists business cards to read 'Flame Assistant' (sorry Brian, couldn't resist 🙂

As for no one askign for a flame on mac... where have you been? Its the most requested thing after SLI and mutilple GPU in Flame. No one ever wanted to learn smoke on mac (well maybe some people but not me) we actually all just wanted (and still want) flame on a mac. This gets us all one step closer to that reality. 

In conclusion, I like the plan, I like the direction and I love the attention flame is getting to put it back on top of the client driven VFX business.

[julioleon]

Great Brian. It could not be more clear! My only concern is the compatibility with flame.

[BKM]

---

@Anonymous wrote:

Also, the biggest expense for post houses will be reprinting their Smoke artists business cards to read 'Flame Assistant' (sorry Brian, couldn't resist 🙂

---

Hey If Mike's on board then surely there must be something wrong.   🙂

And I am now a Flame Editor.  As well as a Smoke Artist.  So I'm good with it all.

[Anonymous]

I don't know Mike - four or five people on an informal website where people speak "mostly" freely is not the same as every major firm on the planet who have more than four licenses installed.

Once you scale flame past two workstations it can be reasonably complex to maintain.

Plus burn

Plus flare

Plus Smoke

Think of the inconsistency of every mac pro workstation in every firm on the planet?

You want to maintain that randomness?

Even for a premium?

You said it yourself - flame & flame assist (although you should exploit the flame premium offering IMHO)

When the prospective user base wants to pay less than 10% to get the same tools?

Doesn't make sense.

& yes I know "think different" & all that.

Well "pay different" & you'll probably "get different"...

[Anonymous]

The big facility owners of course no more want flame on mac than they want smoke on mac. Such an abboration will simply let people know its just software on a computer and not a magical 'system', but in the real world that genie has been out of the bottle since the move to Linux. Flame as an entity is today no more mysterious than a copy of Word, its just trickier to install... and I think thats the point of being a systems customer. I dont have to maintain much at all. I pay my hardware and software subs and I gets same day visits from a guru from Comgraph who says 'thats odd' for a while while typing in a Unix shell. I have lunch with the client and come back to a working system.

Anyone who has ever done any software development knows how hard it is, how much bigger every job is than it first appears and how to get things out on time you have to basically throw money at every problem. Forever. So our industry's race to the bottom has to end somewhere. I for one am grateful to see the brakes going on everywhere. Avid pushing Isis for 135K, cloud subs for 35K, Nuke studio etc all positioning themselves for a fight to the death. To do that you need a warchest and 3500 bucks only buys so many bullets.

Im an end user here like everyone else, I would love flame to be 20K (and on a mac) but the purist in me wants the fastest tools on the planet. I have in the past put my money where my mouth is and bought Onyx's and SGI harwdare up the yazoo but the rewards are there. No one ever lost money on buying a flame until they stopped keeping their platform current or skimped on the artist. As an owner operator I'm essentially a cook in my own kitchen, so I naturally want the best set of knives.

I came into my new company after a year of hell, most of my money spent on heart surgery and other things to keep me alive, so I was very grateful taht I could setup for not much money. But a few years in Im at the point where I have zero faith in Smoke as a developing effects tool purely based on its small user base and too cheap price structure. Its like Avid DS all over again, why did Avid stop selling it? Because no one was buying it.

So it is with SMoke 2013, its between two stools, too limited for the vfx guys, too complex for the editorial guys. AD took some wrong steps in its develoipment but in truth I said very early on the needs of offline editorial and online are massively different and in my opinion irreconcilable with current technology. Until the edit systems can do everything with uncompressed video that an offline system of today can with tiny files then both systems are needed. And 4K doesnt help that become a reality. Nor does s3d or high frame rate. Its the moving goalposts that keep us all on our toes and make this industry such fun.

Pay different and get different is actually the key here. I think if AD get a useable version of cheap editorial smoke and add enough flame features into flame assist then they can have a very positive vertically integrated product range. The leap from smoke to systems will become harder though and sadly from an end user perspective that is actually a good thing, business plans always contain discussions of 'barriers to entry' and cost is a good one. Personally I love the democratisation of the tools and the wide range of people I've consequently met on Area, years of just reading Flame news insulates you from how diverse our industry is and the smoke 2103 attempt at change brought a lot of new opionions and attitudes to the fore which benefits everyone. The business part of me though has to admit I like the idea of there being less companies chasing my clients... (It's kind of a moot point here in Bangkok where theres already about 40 flames anyway...) And lastly the nostalgic in me loves the idea of flame having one more go at being top dog. I'm not sentimental about systems, I dropped Harry and Henry like a hot rock once I saw Flame, but as a very long time user at this point there is something kind and familiar about the Irix toolbox colour coding and its familiar menu structures. I'll readily admit I'm in the last years of my career at this point and though I will undoubtedly get Fusion 7 (THAT'S how to do GPU acceleration) and Nuke studio (very cool) I like to think I'll still retire as a Flame artist. Flame artists are sexy after all.

[jamesgrean]

"Pay different and get different"...

Nuke+Hiero is 10k. Nuke Studio I would expect under 15k. And they don't compromise on toolsets. These programs actually do more than a Flame (except give the op "people skills").

What I want, is an Autodesk product that can compete, really compete, with Nuke Studio. And currently there is not one for under $100k.

SMAC can't do it if you don't have the ability to add sparks, use reels, code your own macros and gizmos, use a render farm. So why is anyone going to buy it?

As someone else said, SMAC is not enough for a VFX artist, and too overwhelming for an editor. It's in no-man's land. Just let me buy "Flame assist" for 10k and be done with it. Protectionism is losing you buyers, me included.