[Tig] Re: Lustre Review

Christopher Noellert cnoellert
Sun Mar 27 23:35:09 BST 2005


On 3/24/05 1:28 AM, "Jack James" <jack at digitalintermediates.org> wrote:

> Thanks to Chris Genereaux for supporting the TIG.
> --

This is going to come across harsh.

> If I may:
> - Lustre 2.5 was in beta at the time I reviewed it.

Hmmm.  2.5 was released at IBC last year.

> - Baselight is indeed a worthy opponent to Lustre, expect a review of v 2.0
> very soon.

If you haven't driven Lustre how can you possibly compare it to anything you
have driven?  Have you driven Baselight or is the intent to write a review
with out having driven that one either?

Just curious here. 

> - I've no hands-on experience of using IQ, so it would be senseless of me to
> make comparisons; however, from trade show demos and quantel's demo dvd I
> believe it still is lacking in some areas, but I'm not going to cast
> aspersions until maybe after nab

Having no hands on experience shouldn't hold you back at this point.

Having driven all three packages they represent a very very common toolset.
Color wise there is nearly nothing possible in a Lustre that cannot be done
in an iQ feature wise.  The inverse I would say is NOT true.

There are three major points of interest.  Workflow, speed and toolset.
Baselight8 is the speed champion, Lustre is pitifully slow compared as is
the iQ. Workflow wise it's entirely situational.  If you've got 3 Infernos
and 6 flames and won't to go into DI than a Lustre is going to make a whole
lot of sense.  If you've got a Northlight scanner and truelight hardware
than of course you're looking with kind eyes at Filmlight.  If you're
looking for a more all in one approach with very little data movement than
the iQ is the champ.

Different strokes for...

As far as toolset is concerned I think for coulor we're talking about more
or less the same stuff.  Outside of colour correction there's a few places
where things are better are worse.  For example how do you calibrate a
Lustre or an iQ?  My solution is Truelight, which of course is built into
Baselight.  How are you going to get your material conformed and on to the
system?  IQ is a full fledged editor.... Not just a conforming timeline.
How will you handle revisions and reedits.  Bselight and iQ have great tools
for doing revisions.  How will your assistants work?  Lustre and Baselight
have assistant workstations which can tie into the main station over the
network.  

There's so damn much more than colour that needs to be apprised when doing a
review.  Someone who writes for an expert site for DI would know what issues
are important to the industry.

> - I'm not a fan of hardware-based solutions since I had bad experiences with
> the specter- I honestly believe that software solutions will prevail in the
> 2k world.


I get the impression you don't know so much about the 2k world.  But maybe
it's just because you're not giving us your best side.

The lines between hardware and software are blurring.  For Lustre, I would
imagine the future is cluster based computing on multiple GPUS all tied
together with some kind inifiband nas.  For Baselight8 it's the speedfx
platform which is essentially what I believe discreet will rip off for the
aforementioned setup.  Nucuda and everyone else will go to multiple GPUs in
some form or another.

I guess my point here is that "software solutions" are still only as good as
the "hardware solutions" they're devised to run on.  There are limitations
in software just as there is in today's softprogramable hardware GPUS used
in DaVinci's and Pandora machines.

> -I'm a discreet fan only in terms of the UI. For my money, Combustion is one
> of the best Price/Performance systems available on any DI platform.

Combustion is not a DI product in any way shape or form except for maybe an
assistant station for cutting mattes for a color corrector or doing effects
shots.  Because it can be used in a DI pipeline does not make it a DI
platform.  

Then I can say, I like Shake.  It's a fantastic price performance software
for DI.  It hurt me to write that sentence because it's so damn stupid.

> - The only systems manufacturer that pays me at the moment is Thomson to
> consult on the new Bones system, and although I fervently believe that Bones
> is a remarkable system (unlike specter/phantom transfer engine), I will not
> include it in any reviews for this very reason.

I would prefer for you to write about something that you know about rather
than something you're speculating over.

> - My comment about graders not needing to edit is valid- colourists should
> not need to have editing skills. The point is that systems such as Lustre
> desperately need to accommodate the fact that people will be using it to
> edit, and that marketing an all-in-one solution is futile if it ultimately
> only caters to the needs of the colourist, and does not consider completely
> the requirements of the conform assistants, digital asset managers and
> telecine assistants. Someone used the term "specialisation", and I believe
> this is key.

To do anything in lustre you need to create an edit by placing clips from
the disk on to the timeline.  To do anything more meaningful than that
you're going to need at looking to integrate some sort of conforming based
on an edl or by having an Inferno, Fire, Flame or Smoke hanging around on
the network that you can access the edit in question.  Regardless editing is
part of what colourists do even today.  You won't use Lustre or baselight to
edit per say, but you might use and edl or edit to set your events in a
list, or to conform a series of edits from disks.  This is where the
difference lies between what you're talking about and what Steve Shaw is
discussing.  Editing in Lustre is like trying to hotsplice with a brick.
The same holds true for Baselight.  Although in my opinion the multi-layer
timeline in Baselight makes revision a ton easier to deal with then the
single layer Lustre timeline.

Still in the more dedicated colour products we're talking about a conforming
timeline - very much similar to what Inferno and Flame were running in the
early days.

> - To address what Dave Picket asked, as far as I am aware- Lustre cannot
> directly control a telecine, it is a software-based approach that only works
> with digital images, and has no option for live signal (or tape-to-tape)
> grading. In order to incorporate a Lustre system, you must already have a
> system in place to allow for capture/digitisation, be it a film scanner or
> SD/HD video input. Incidentally, this is also the case with Baselight,
> Assimilate Inc's Scratch, and Quantel's IQ.

I think it was pointed out but just to say it again, Lustre will be able to
drive anything that supports a 422 control or emulation.  This is the same
with all the products you mention.

Chris

--
Chris Noellert
Frithiof Film to Video. AB
Sturegatan 58 
114 36  Stockholm

Tel: +46 8 545 678 78
Fax: +46 8 545 678 79
Mob: +46 735 32 00 03
AIM: cmnoellert
www.frithiof.se







More information about the Tig mailing list