Scratch vs Pablo

[Darren Orange]

Hey Guys....It's ones and zeros here, either it handles it or it doesn't, there is nothing subjective about this.

This argument is almost like comparing HDMI cables, HDMI is a digital cable, it either works or it does not.

"Precise" sounds like marketing propaganda.

 

[Allan Stallard]

I have zero issues with discussing products and pros/cons. I just want to know what the heck a "precise" color space is.



Lucas

I figured you wouldn't Luki, I was actually referring to the other parties involved.

I'm not sure I feel comfortable with exact quotes, I would say that something along the lines that was quoted by the original poster seems familiar. I too would like to know what these actually mean. Using Color for instance I find some of the curves very easy and "precise" to use. Desaturating a specific color range is pretty straight forwards. On Flame/Smoke I find selecting color ranges a bit more involved as I am forced to use the Color Warper or creating a comp. To me the precision of that particular curve was better in Color. Yeah I know that precision in can mean 2 different things, the color space/bit depth ect. and/or the actual tool that I am using. The "tool" in Color to desat a specific color range was IMO more precise than the similar tool in Flame/Smoke.

(Yes I know Flame/Smoke are not not color grading specific apps, and they have lightspeed hardware differences. Just making clarifications and comparisons)

So saying that, is it possible that the original quote was referring to tools sets more so than bits bytes and tech specs?

Also the original quote was from a user/operator not someone from Pablo. What if the quote came from someone who has both Scratch and Pablo? Is it not fair to inquire how this opinion (as that's what it is at this moment) came to be? I wish someone with this opinion would join in and give their 2 cents. That is the problem here. We are missing part of the equation for a decent discussion...

Also I would hope that it was obvious that just because a product is more expensive doesn't mean it's better, but realistically c'mon we are talking about a system that is 5 times more expensive than the other. This is NOT a direct comparison. This discussion is happening because someone said Scratch is not the BEST color grading app. Couldn't this be true? Second best? Third? Whatever, is Luki actually saying Scratch is the best? Other users? I don't think so, what I hear so often is "best bang for the buck". So taking the "buck" out of the equations what are some differences>?

 

[Florian Stadler]

Plaster city recommends red conform in scratch and color work in pablo for the highest quality filmout. They hinted at problems with skin tones resulting from flawed color fundamentals with Scratch. This is strictly for a film out as they recommend scratch for a video finish. It might have something to do how scratch handles luts but this is only a guess on my part.

 

[Lucas Wilson]

So taking the "buck" out of the equations what are some differences>?

I'm sorry, but SCRATCH is the best in all categories at all times. I thought I'd made that clear?

:tongue:

Lucas

 

[Allan Stallard]

I'm sorry, but SCRATCH is the best in all categories at all times. I thought I'd made that clear?

:tongue:

Lucas

I cannot deny your enthusiasm for the product! But which is better: The White Stripes or The Raconteurs?

 

[M Most]

Plaster city recommends red conform in scratch and color work in pablo for the highest quality filmout. They hinted at problems with skin tones resulting from flawed color fundamentals with Scratch. This is strictly for a film out as they recommend scratch for a video finish. It might have something to do how scratch handles luts but this is only a guess on my part.

A LUT is a LUT. Not speaking for Plaster City, but my suspicion is that their enthusiasm for Pablo stems primarily from its ability to work in 4K in real time, with true 4K playout (even though I believe they're using a 2K projector), as well as its ability to output other deliverables without a large amount of additional rendering. This does, of course, currently require a conversion to DPX files, but it does allow direct viewing of manipulation of 4K frames. Pablo does have a very deep toolset, which is nice, but does not in itself translate to "additional quality." Remember that time is money, so unless you're a large studio picture that's willing to spend quite a bit on the DI - usually not the case with current Red productions - the ability to "do more" doesn't really help you in the sense that you have a limited amount of time to complete the work.

 

[John Tissavary]

I agree with Mike M (this is getting to be a disturbing trend .

I've had numerous Scratch & Pablo discussions with folks from Plaster City, and although I would never presume to speak for them, I have not heard any comments from Michael Cioni, their chief technology guy, regarding color issues in Scratch. Their preference for Pablo appears to be rooted in its 4k realtime uncompressed capabilities. I also imagine the extended featureset of Pablo vs. Scratch is another reason they chose Pablo for hero work.

Scratch has areas of weakness vs. its more expensive competitors: keying, masks, tracking, and being restricted to a single timeline, are a few examples. For color, my experience is that it's as accurate as any other package out there. None of them are slouches in the color department - it's a well developed, thoroughly researched area of computer graphic imaging.

I'll really need someone to break down for me, empirically & mathematically, how Pablo is more 'color accurate' than Scratch, because at this point in time, based on my experience, I do not find this to be the case at all, and the heresay and supposition is really making the rounds on this board lately.

The skintone argument seems really popular lately. There's no magic button in Pablo or Baselight or Scratch for getting good skintones - IMHO it's the ability of the colorist that puts the skintones where the client wants while still being able to work the surrounding pixels.

If someone is more comfortable working with a Pablo toolset, then they'll probably get better skintones in Pablo over Scratch, but to blame it on color inaccuracies does seem rather suspicious to me.

JT

 

[Mark L. Pederson]

I'll really need someone to break down for me, empirically & mathematically, how Pablo is more 'color accurate' than Scratch,

pure fiction

The skintone argument seems really popular lately. There's no magic button in Pablo or Baselight or Scratch for getting good skintones - IMHO it's the ability of the colorist that puts the skintones where the client wants while still being able to work the surrounding pixels.

If someone is more comfortable working with a Pablo toolset, then they'll probably get better skintones in Pablo over Scratch, but to blame it on color inaccuracies does seem rather suspicious to me.

JT

post of the day - are we done here yet?

 

[Jeff Brue]

I've played with nucoda, baselight, pablo and scratch. I'm going to have to go with lunacie. It's really about the talent of the colorist. I've seen some amazing things happen in scratch with a full set of tangents properly mapped out.

I really do think most of it has to do with artist experience, a program like Maya can easily take several years to learn and become a pro at (and thats if you have the capability to even do it). Color correction is the same thing an art/science, just normally with much higher budgets attached. Also with a colorist with only a year of experiance on a system often charging the same amount as a person with three.

Confusing waters and it's often easier to market the toolset then a colorists ability to nail the look in 3 seconds, with only a few spins of the wheels. But the reality is thats what should be expected and all of the above platforms are more than capable of doing just that.

 

[Todd Jaspers]

I think in the end Pablo and Scratch can have the same result, but the Pablo has the benefit of a bit more of R&D and development. Quantel uses propriety MME boards for rendering and coloring, they have a great toolset with Qcolor. The HSL keyer is really great. Quantel does really accurate color. The color space conversions are so f*cking accurate that you could do YUV to RGB conversions for 2 years on the same image and not be able to really tell any color change.

But if you purchase a fully setup Pablo iQ you are buying a 600k color grader. You could buy three scratches with that and have some money left over. I would say the Pablo is a more competent color grader, but you have to pay the extra price to get that extra 10% umph!

If you know you are going to get as much work as plaster city than thats not an issue, for the DIY DI facility Scratch is probably more affordable. DIY can easily become DYI if you are financing a 600k system.

-T

 

[M Most]

But if you purchase a fully setup Pablo iQ you are buying a 600k color grader. You could buy three scratches with that and have some money left over. I would say the Pablo is a more competent color grader...

I would say it's more efficient, has a deeper feature set (in part because it's basically a full editing/conforming/effects/finishing system - basically a full Quantel IQ - coupled with a color corrector), and has better keyers. All of which I'd expect for the price. However, competence, or lack of it, resides in the person running the equipment, not the equipment itself. As does talent, and ultimately, the responsibility for the result.

 

[Todd Jaspers]

Yes mmost, I would rather Grade on an iQ Pablo than any other system out there... BUT!! Be honest, its Compositing features suck. I feel like I am on After Effects 3.1, with a cheesy 3D engine. I would rather do effects and compositing on any other thing out there, Flame, Nuke and Shake.. Even After Effects!!!

For DI/Color Grading/Finishing the iQ is amazing... I just can't believe with all the R&D at Quantel they can't rebuild a better compositing engine.

 

[M Most]

For DI/Color Grading/Finishing the iQ is amazing... I just can't believe with all the R&D at Quantel they can't rebuild a better compositing engine.

Well, Quantel has always done things their own way. And if you don't really understand the Quantel way, it can be pretty baffling. In my earlier days as a visual effects supervisor, I worked with a lot of very talented Henry operators. They could turn out great work, but unlike any other system I've ever sat behind - and that would include Flame and many desktop programs - I could never, ever figure out how they were doing it. The menus were goofy, the interface was cryptic, just when I thought I understood where they were going they would go someplace else - I found myself totally confused by what I was looking at.

Like I said, you have to either "grok" the Quantel way, or not. Although in fairness, the iQ line is far less baffling than Henry ever was. But it's still baffling.

 

[Todd Jaspers]

I would say 5 years ago this were true, but now with the real 3D cameras in NUKE, IFF and even AE. Taking a 2D Background plate with 3D camera data extracted from it and doing a real 3D composite is impossible in Quantel. Also with lack of HDRI support, and the fact you can't just click a button to see the R, G or B channel is silly, if not just plain stupid. The iQ will keep people with its great color grading, but as soon as someone has a high end station with 3D compositing and Good color grading in one app, I am there.. Autodesk has kept everything separate on purpose to make more money, but someday they might change that.

 

[M Most]

The iQ will keep people with its great color grading, but as soon as someone has a high end station with 3D compositing and Good color grading in one app, I am there..

Sounds like you want a Smoke. Or perhaps a Flame. Or perhaps Smoke and Lustre on the same workstation (there are deals to be had for that combination......).

 

[Todd Jaspers]

Don't know what I want. Don't think it exists. I do want to see the SGO Mistika in Action. But no demos exist in the USA.

 

[M Most]

Don't know what I want. Don't think it exists. I do want to see the SGO Mistika in Action. But no demos exist in the USA.

Actually, now that I think about it, what you might want is a Piranha system, albeit one with Red support. That, of course, won't happen until there is a Linux SDK release. Perhaps a more evolved version of Eyeon Generation would also be interesting.

 

[Lucas Wilson]

Man, I really don't like threads with titles of "ProductX vs. ProductY."

I worked for Quantel for a brief period and have a tremendous amount of respect for Mark, Martin, Simon, Steve, Damon, George, Eric, and many many other names I'm leaving out. Quantel has been an incredibly innovative company in production and post-production, and continues to innovate. I mean - have you *seen* the Neo? It's absolutely amazing. Overkill for a lot of people, perhaps... But Quantel has always specialized in the ultra high-end, where customers do not mind paying a lot of money for dedicated systems, because it is worth it to them and their client base.

And the fact that so many of their employees have been with Quantel for so long is a testament to what kind of a company they are and how they treat their people.

The eQ, iQ, and Pablo are incredible systems that do great work. I also think SCRATCH is a great system for many of the same reasons, and for a lot of different reasons.

Different tools for different needs.

Best,

Lucas
-----
ASSIMILATE, inc.
LA, CA, USA

 

[Michael Cioni]

Modern Post Production

Modern Post Production

For those of you who do not know us, PlasterCITY is a DI facility that specializes in data-centric post production. Thanks to Assimilate, we've been able to expand our DI services to include projects shot on the Red One.

It is been our experience that in modern post production, no one tool is capable of covering all the bases of DI. As client expectations continue to rise, so does the need to invent workflows and modify toolsets to close the expectancy gap. Couple that with the need to conserve post production costs and shorten deliverables schedules and you will find that it takes a lot of different tools and even more talent to navigate changing waters.

The entire post production community has been charged with the task to retire old models and replace them with an evolved infrastructure. -We find this evolution is as much in our minds as it is in our tools. Some entities thrive on this progression, others choose to avoid it and some just like to debate it.

Regardless of your position on the matter, RedUser is an encouraging place to experience progressive attitudes and challenging discussions. But with regard to Scratch and Pablo specifically, PlasterCITY successfully uses both systems, simultaneously and most often on the same jobs. Without Scratch, none of our Red work would see the light of day. As our clientele has come to experience, on the right job, a good artist and a Scratch are able to exceed everyone's expectations.
Michael

 

[gbalaji]

congrats michael and team for sucessfull color timing of india's first feature film "achchamundu achchamundu", which was shot with red one using scratch and pablo in combination. Would like to know more about the workflow. I worked as post consultant (India).