open ended discussion regarding Mac Pro future and post-production hardware/software

[ebobster]

FCP X and the Future of Mac Video post-production
I'm an conscientious and savvy mac enthusiast. I love and hate apple for so many reasons. I have a 6' x 4' poster on my wall of this photo: http://osxdaily.com/wp-content/uploa...ibm-finger.jpg. The prior sentences serve to accredit myself in the eyes of likeminded apple lemmings interested in intelligent discussion.

Now for my confusing question(s):

I advise people on apple hardware and software for a living among various other things and am currently trying to help a small production company with a significant hardware purchase and have come to realize that the status quo in apple's video editing products is about to change markedly.

I am by no means expert in this field which is why i seek advice.

FCP X appears to be moribund software. Xsan already is. Xraid too. At the time these product lines were relevant and supported unwaveringly and amounted in my eyes to effective albeit exspensive solutions for video editing. Then came RED cameras and native 5K resolution and the demand for faster everything. If you want to survive in this industry you need performance but apple can no longer offer a reasonable setup for a small studio in need of a file server for 6 desktops. Please object if this is not the case, but the damn FibreSAN card alone would dent my budget noticably.
So I think the best approach to be very cautious and recommend a highly upgradable and expandable setup. That's ideal, but in reality, i feel otherwise since such a setup would be in the high $20K+ range.

The studio seems intent on moving to CS6 and I think thats wise since Adobe has a devotion to the professional media user base. My main interest is in hearing the thoughts you might have about what you think professional video editing is going to look like on a mac in 5-10 years and whether or not a rock solid file server and some highly tuned iMacs will be useful down the road. I'm aware that the mac pro line has been promised a noticeable revamping in the months, and barring an ****** move by apple that solders every god damn piece of hardware to the mac pro, that would be the more cost effective route to go as you can start small and go pretty big with those machines.
Is that the prudent and wise route to go for professional grade performance that won't be choking on the video file of the future that is presumably larger?

How much can one completely tuned up model year iMac handle at once? I ask solely because I know what they can theoretically do but the temperature of their 2010 iMacs at capacity is borderline unsafe. My temp gun showed 165F on one of their machines. This leads me to believe that they'll run longer and more efficiently at much less than their capacity. Common sense, yes. Common practice, No.


So just tell me what you would do if you were a indie-type studio looking to get the most bang for your buck on a mac. Or if you would scrap this idea altogether and build Hackint0sh boxes from scratch with a hardware selection geared solely for video editing.

Much obliged,

Harry J

 

[d-m-a-x]

i think it's 50-50 at this point, maybe more like 40-60 for video people. lots of people have jumped ship to Windows and Hackintosh. A significant amount of people are still waiting for a new Mac Pro. I opened up Fcpx last week and it was unusable. I did a little research and found that with third-party plug-ins you could restore its functionality. But have no idea how this would interact with high-end hardware, most of my videos are standard def (I'm a Photoshop jockey). I have the luxury of waiting until Apple figures out what it wants to do. if I were a video guy, I would probably jumped to avid and windows by now

 

[JesterJJZ]

I'm one of those guys hanging on to Apple by a thread, hoping they get their $#!+ together. Still rockin' a 2006 MacPro with upgraded processors. I got a RED this year, and while it's a bit slow, I've learned how to jump through the hoops to make it work between FCP and CS6. I guess I'm at the point where I'm giving Apple one last chance with this magical MacPro revision we're expecting in 2013. I had $5k+ ready to blow on a new system in June, but after that slap in the face I shut my wallet and got new processors for $100 on ebay.

If I was running Apple, or at least the "Pro" division, these would be the first things I would do.

1. Rebrand FCPX what it really is, iMovePro (effective immediately with announcement of following)

2. Resume work on FCP8, same (or updated) interface, same functionality, move it to 64bit, multithreaded, etc...

3. Haul a HP820 Red Edition in to Apple and dare them to do better.

4. Bring back the 17" MBP.

 

[PaulD-UK]

FCP X appears to be moribund software.

I opened up Fcpx last week and it was unusable. ...waiting until Apple figures out what it wants to do.

Hi
Apple has 'figured out' what it wants to do for the next 10 years or so, Pro-apps-wise. Think different...

FCP X is far from moribund, but its very much work-in-progress. Apple's FCP X web page has a Later This Year section, and there will be continuing enhancements.

Its addressing the video editing needs of a 'data-rich' future. That may well not fit your current needs.
So be it.

Apple's current and future software is taking full advantage of Intel's AVX instruction-set enhancements introduced in its Sandy Bridge/Ivy Bridge CPUs.
That's why the current top end iMac or rMBP running FCP X and Motion 5 runs rings around the older Westmere Mac Pros - even the sizzle-core ones.

With Haswell next year Intel are introducing AVX2, so for resource-intensive video editing (Red etc) there is likely even more obsolescence in the immediate future. Maybe that's why "later in 2013" - T. Cook, has a significance?

How soon Adobe will recode their Creative Suite to be optimised for AVX/AVX2 is their business. If they delay they maybe risk being left behind?
Like Adobe's existing software it may well work better under Windows?

So CS6 may currently be the way to go for an indie edit/post-production business.
But Apple feel they have identified a strategic direction for their view of the industry's future needs. That may well not fit your current needs.
So be it.

 

[Pressure]

I think people should drop the notion of Final Cut Pro X being iMovie Pro.

It clearly has features even Final Cut Studio didn't have (while still lacking features) and many that iMovie can only dream of.

Have faith in the platform, it will evolve. It is already heaps better compared to when it was released.

Also, as a semi-professional (hell, I only do national TV commercials) to other "semi"-professionals, if you need a new computer you buy one. You don't moan about the one you got now or stick with it, if you need a new one.

 

[Small White Car]

I honestly don't know if this is a good solution anymore, but as of 3 years ago my last job went with Final Cut 6 and an Avid server. (With 5 Mac Pro stations at our desks.) The Avid server was a better solution than anything else for us even though we didn't use Avid to edit.

I suspect that Premiere could work equally well with it.

But I'm not an engineer, you'd really want to talk to an AV shop that sets up Avid systems to find out if that's at all appropriate for your client. I'm just sharing my story since that worked for us.

1. Rebrand FCPX what it really is, iMovePro (effective immediately with announcement of following)

2. Resume work on FCP8, same (or updated) interface, same functionality, move it to 64bit, multithreaded, etc...

I disagree with this move. If they wanted to bring back the option of the old timeline, they should expand FCX into the same sort of beast DVD Studio Pro was: Something that has 'simple' or 'advanced' window setups that you can switch between depending on what you want to see.

It's likely that a magnetic timeline or a track timeline couldn't be converted back and forth, but that's fine. Just ask me which kind I want when starting one and let me have both kinds side by side.

I agree that I'd love to see a "Final Cut 8," but FCX has some neat things in it. I'd rather not see them get tossed aside completely.

I can safely say that Aperture does everything that iPhoto does, plus more. That's an easy choice. I wouldn't want a situation where one Final Cut does different things than another one, but one is not a sub-set of the other.

 

[theSeb]

I opened up Fcpx last week and it was unusable. I did a little research and found that with third-party plug-ins you could restore its functionality.

I really don't understand comments like these, since I can use it. What exactly makes it unusable?

 

[JesterJJZ]

I really don't understand comments like these, since I can use it. What exactly makes it unusable?

Not everyone has the same needs. I need tracks for example, lots of them. It's not up to the NLE to decide when I can have more tracks on my timeline. When I want to start my edit with 20 blank tracks I should have that option. Just one of the annoyances I noticed in the initial version. Maybe they changed that...

 

[theSeb]

Not everyone has the same needs. I need tracks for example, lots of them. It's not up to the NLE to decide when I can have more tracks on my timeline. When I want to start my edit with 20 blank tracks I should have that option. Just one of the annoyances I noticed in the initial version. Maybe they changed that...

So basically it all comes down to the magnetic timeline vs track based interface?

Again, I fail to understand this gripe since you have practically the same thing... I don't understand how a magnetic timeline could possibly limit one's creativity.

http://fcpx.tv/Pages/tracks.html

 

[Small White Car]

So basically it all comes down to the magnetic timeline vs track based interface?

Again, I fail to understand this gripe since you have practically the same thing... I don't understand how a magnetic timeline could possibly limit one's creativity.

It's not about creativity, it's about working with others.

When your FX guy wants the chroma key shots and the background plates exported to him separately and the colorist wants the original files with the final compositions but NOT the plates...well that's a heck of lot easier to manage if you've kept all of those things on different tracks from the start.

Final Cut X is fantastic if you're working alone. It gets a bit trickier when you're in an office with a whole team.

I'm not saying it's impossible. I'm just saying it's trickier. Notice in your graphic that in the magnetic timeline they have blue, pink, green, red, and orange all on the bottom. That's a problem for what I'm talking about, and that IS how magnetic timelines tend to end up.

 

[theSeb]

It's not about creativity, it's about working with others.

When your FX guy wants the chroma key shots and the background plates exported to him separately and the colorist wants the original files with the final compositions but NOT the plates...well that's a heck of lot easier to manage if you've kept all of those things on different tracks from the start.

Final Cut X is fantastic if you're working alone. It gets a bit trickier when you're in an office with a whole team.

I'm not saying it's impossible. I'm just saying it's trickier. Notice in your graphic that in the magnetic timeline they have blue, pink, green, red, and orange all on the bottom. That's a problem for what I'm talking about, and that IS how magnetic timelines tend to end up.

Hmmm. I see, now. That makes more sense on why a track based UI is useful. The funny thing is that I've read a lot of gripes about this and, frankly, you're the first person to explain the problem so succinctly.

 

[Small White Car]

Hmmm. I see, now. That makes more sense on why a track based UI is useful. The funny thing is that I've read a lot of gripes about this and, frankly, you're the first person to explain the problem so succinctly.

It's worth noting that I do think FCX is cool and I do recommend it to certain people.

And at the same time, I didn't suggest we use it here at work.

It's not so much about FCX being 'bad' as it is 'not enough.' Those are different things. I think FCX is a LOT like the first iPhone. Remember that thing...the one without even an app store?

Imagine if that phone had come out 3 years later. Things would not have gone so well for the iPhone. That's kind of what FCX is. It's not bad, but it's just not enough.

I'm hoping it improves as much as the iPhone did, but so far that's just a hope, not a reality.

 

[CASLondon]

the Hackintosh route is tempting price-wise, but daunting on the support/keep it from breaking/drivers/software/hardware integration route.

I don't see any reason not to stay with Mac Pros, for a small studio even these dated machines would be fine, optimized for production. Waiting for Thunderbolt is not required for production, other than in the laptop and mobile production sphere. There, I would suggest non-retina MBP and Pegasus Thunderbolt Raids.

Even if the beefier iMacs are fine for running current generation software bundles, the increased robustness and upgrade paths available mean I would stay Mac Pro, if money a problem then stay one generation back, second hand, etc. And upgrade as needed.

Here's some tips, though from a photographer's perspective lots of good advice...

http://macperformanceguide.com/

If you disagree, lots of Avid editors are using the HP Z800/820 work stations.

Software-wise, if you are making broadcast tv or long documentaries, high shooting ratios, multiple editor projects, etc, I'd run Avid plus Aftereffects or, especially for short form, CSS 6.

FCX is fine for one man band, web video, same day edit jobs, wedding videos, etc.

Media management is most solid in Avid world, in my opinion

 

[JesterJJZ]

So basically it all comes down to the magnetic timeline vs track based interface?

Again, I fail to understand this gripe since you have practically the same thing... I don't understand how a magnetic timeline could possibly limit one's creativity.

http://fcpx.tv/Pages/tracks.html

Adding to the points above.

I use empty or blank tracks a lot. Either as a means to separate parts of my edit or as scratch tracks to temporarily place things on. Once you remove everything off a track it disappears...arrgghh! It's like it tries to anticipate what I'm doing and is wrong 100% of the time.

Having tracks and physical track lines across the screen makes it easier to see where stuff is as well. I like to place certain things on certain tracks and I'm not allowed the freedom to edit the way I like in FCPX.

When you're editing a project for a long time, you can eventually make your way around the edit just by looking how everything is laid out. You memorize what your edits look like almost like a map. If the timeline constantly changes and track appear, disappear, or move around, you just can't "learn" your own edit.

It's a just a lot of little things that make it "unusable" for me.

BTW can you "save and save as" your project yet? lol! Or are they still calling them "events"?

 

[GoreVidal]

If I was running Apple, or at least the "Pro" division, these would be the first things I would do.

1. Rebrand FCPX what it really is, iMovePro (effective immediately with announcement of following)

2. Resume work on FCP8, same (or updated) interface, same functionality, move it to 64bit, multithreaded, etc...

3. Haul a HP820 Red Edition in to Apple and dare them to do better.

4. Bring back the 17" MBP.

Yup, ever single one of these points.

----------

It's not about creativity, it's about working with others.

When your FX guy wants the chroma key shots and the background plates exported to him separately and the colorist wants the original files with the final compositions but NOT the plates...well that's a heck of lot easier to manage if you've kept all of those things on different tracks from the start.

Final Cut X is fantastic if you're working alone. It gets a bit trickier when you're in an office with a whole team.

I'm not saying it's impossible. I'm just saying it's trickier. Notice in your graphic that in the magnetic timeline they have blue, pink, green, red, and orange all on the bottom. That's a problem for what I'm talking about, and that IS how magnetic timelines tend to end up.

This exactly as well. Sheesh, it's like amateur "editors" like it when the company making the NLE dictates their workflow. That's not how it works in this industry.

 

[derbothaus]

XSAN was just revved and added some key features. Features that make it easier to move to Linux and StorNext but hey, something. And of course completely silent. Now it's just too bad 10.8 clients have like 1 Xserve they support and that can be used and chances are you don't have it
If something doesn't change it will be time to move on. No current GPU's, outsourced RAID chassis, only Mac Mini and Desktop server options with no LOM, redundant power, etc. No upgraded Fibre card in years (or price drop), no desktops with Thunderbolt to take over for the lack of what I listed above. It is a minefield of poo.

 

[d-m-a-x]

I really don't understand comments like these, since I can use it. What exactly makes it unusable?

All of the precise video and audio filters got replaced with things like "sepiatone", and "8 mm". Where did the broadcast safe adjustments go?

 

[linuxcooldude]

All of the precise video and audio filters got replaced with things like "sepiatone", and "8 mm". Where did the broadcast safe adjustments go?

There is a video filter called "Broadcast safe" is that one your talking about?

 

[goMac]

2. Resume work on FCP8, same (or updated) interface, same functionality, move it to 64bit, multithreaded, etc...

Such a thing is impossible. Most everything FCP7 is based on was cut by the operating system teams in the 64bit transition.

The FCP7 codebase is an unusable dead end because of this. Sure, maybe Apple could have handled things better, but a full rewrite was completely unavoidable. FCP7 might as well have been written in Greek. The two main APIs it was based on (Carbon and QuickTime) no longer exist and don't have any drop in replacements.

 

[GoreVidal]

Such a thing is impossible. Most everything FCP7 is based on was cut by the operating system teams in the 64bit transition.

The FCP7 codebase is an unusable dead end because of this. Sure, maybe Apple could have handled things better, but a full rewrite was completely unavoidable. FCP7 might as well have been written in Greek. The two main APIs it was based on (Carbon and QuickTime) no longer exist and don't have any drop in replacements.

Too bad the ability to recreate the interface, features, and usability using some kind of programming language has been lost forever since FCP X hit.

 

[goMac]

Too bad the ability to recreate the interface, features, and usability using some kind of programming language has been lost forever since FCP X hit.

Heh. Yeah, they could have tried to re-create as much of the UI as possible. But some features that were dependent on QuickTime are gone for good. And if they had tried to recreate everything we would have been waiting for years more. The noise about not having 64 bit was already reaching a peak right before FCPX was announced.

It would also still be impossible to have old projects look the same due to the different pipeline. So project import would still be difficult.

 

[DJenkins]

But Apple feel they have identified a strategic direction for their view of the industry's future needs. That may well not fit your current needs.

Fair enough you've acknowledged it's not going to fit everyone's needs, but I think their 'strategic identification' is more Apple dictating how they think the industry should work, not listening enough to what people actually want.

I think people should drop the notion of Final Cut Pro X being iMovie Pro.

It clearly has features even Final Cut Studio didn't have (while still lacking features) and many that iMovie can only dream of.

None of what happened in FCX really surprises me, Apple's philosophy is obviously to break things down into their most simple form... however a lot of pros need a program that is going to be flexible enough to adapt to their unique workflows as you guys have just said. Not just to be told "our way is the only and best way to do things". Maybe it will get there in the end but it looks like a slow and painful path just to get back functionality we already had.

As far as ideas for a small to medium production house, it is a bit of a tricky spot at this point in time. This is due to the big changes around Apple's professional products, but the industry is still very focused around the Mac specific ProRes codec. This pretty much used to be one universally accepted system but now many are looking at other options.

Avid is a great system and really shines with a Unity network system. But it's the sort of thing you really have to go all out on to see it work to it's full potential, and this means it gets expensive! Especially if you get to the point of having all Avid certified workstations.
Good points - you can have both mac & PC workstations attached. So you can have some tricked out HPs with the latest processors and GFX cards, and maybe a few mac pros if you desperately need to edit/output ProRes from Final Cut.

The cheapest way would be to get the Adobe Suite and use Premiere, with some fast DAS units attached to a few workstations.

As a base I would have as many of the workstations mac pros or PC towers as possible, and then only if budget is tight get some iMacs to pick up the remaining workload. An 8/12/16 core machine with up to 4x RAM and decent GPU will absolutely smash the top of the line iMac when it comes to video production. Especially if these machines need a blackmagic or AJA video card, there is just no other option.
I know you can get PCI expansion boxes for iMacs but I think people are pinning a lot of hope on that technology around a machine which really isn't ideal for handling it.
If you were a photographer or working as a single user on regular video footage, absolutely get an iMac with thunderbolt raid attached. But as soon as you need to function in a multi user environment, sharing heavy video media, expect to see that system hit it's limits.

I would recommend as soon as you need over 3 workstations to invest in a centralised storage system/server. There is very high outlay cost but a few years down the track you will never have looked back. If you need to add another workstation, go for it. The storage system is there and waiting for as many users as you need. There are guys talking about this in another thread here but I haven't read through it all yet.

Having workstations with individual storage attached becomes an absolute nightmare to share and backup, especially if you start expanding and adding users. What was a cheap solution for 2 machines suddenly becomes just as expensive as a true NAS as you expand to several, and you are stuck with a system where users can't share media. You will waste time copying files between machines and having duplicate media all over the place.

So back to the original question...where does Apple fit into all of this?

Well disappointingly they have slipped out of the production pipeline in a few areas.
They do not offer any complete dedicated server solutions anymore.
Their workstations will only be bought out of necessity to stay with the OSX system, not because they are the best.
Their NLE will not be the heart of the production workflow for many places as it used to be.

Wow I'm struggling to find any good points!!! We all love OSX?
I think they are simplifying themselves out of the market!

 

[deconstruct60]

When your FX guy wants the chroma key shots and the background plates exported to him separately and the colorist wants the original files with the final compositions but NOT the plates...well that's a heck of lot easier to manage if you've kept all of those things on different tracks from the start.

If folks were using tracks to implicitly label content as being of some type then perhaps Apple just has to weave this into the metadata/attribute labeling/tracking system the FCPX employs.

So at export you could say "only export objects with property XXXXX, YYYY , and ZZZZZ into the exported data" .

Actually overtly labeling the object is better for long term maintenance of the project.

Adding to the points above.

I use empty or blank tracks a lot. Either as a means to separate parts of my edit or as scratch tracks to temporarily place things on. Once you remove everything off a track it disappears...arrgghh!

Empty/Blank tracks are an implicit encoding. Drop someone not trained in this implicit encoding into the project and they aren't going to know what they connote.

This is somewhat akin to the debate that sometimes goes on in the computer programming world on commenting code and using descriptive
variable names.

It's like it tries to anticipate what I'm doing and is wrong 100% of the time.

More likely it doesn't really think that folks will being implicitly organizing this way with this mechanism. It is likely thinking you are not trying to do something isn't a incremental step in constructing the final solution. (instead of a experimental subproject that is simply trying something out to maybe throw away).

When you're editing a project for a long time, you can eventually make your way around the edit just by looking how everything is laid out.

To some extent that is because you made it. Come back 5-6 years later when your style is somewhat different and it won't be as easy to read.

FCPX appears to put a higher emphasis in grouping across object types than in grouping simillars together in individual tracks with an implicit grouping by timecode across objects.