|
...and thinking about comparisons... |
 |
courtesy of the AE list e-mails of Dik Baskerville
--------------------------------------------------------------------------------------------------------
Message-ID: <005801c1eb29$53e88b60$303d03d2@KINESISADMIN.local>
From: "Kinesis-Eidolon" <kinesis@kinesis-eidolon.com>
Subject: Re: [AE] Which plugs to get?
Date: Wed, 24 Apr 2002 08:43:47 +0800
Gary wrote:
> anyone have the latest count of just how many plugs there are out there
> that will work in AE? I'm guessing around 300-400+?
Not quite.
Here are some stats from the K-tabase. These are:
- Mac only (sorry)
- Filters written to the AE Spec only (viz. not those PS filters which work under AE) - in no way guaranteed to be accurate!!! - does include Plug-in Galaxy (no favouritism here)
Total filters [including Beta versions which I know about - mainly Tinderbox 3 (as posted on The Foundry's site) and Useful Things - these should be released by End of May] which will run under AE in either OS-9 or OS-X
=1095
(note that this figure includes those which are no longer available like Kaleidafex Matte Pack and sets which have been superceded by different incarnations - e.g. Cult FX Vol 1 and Evolution - but which still work in their own right in OS-9)
Total filters which will still be available for sale by end of May:
=951
Total filters available now:
=847
Total filters available for OS-X now:
= 537
If you include announced Betas:
= 634
Of which there are 250 x 16-bit filters
Now I've cheated slightly in that I've included "commercial-level" Useful Things scripts in this: effectively these are like a conventional plug-in. In fact some of these are aimed more at showing capabilities of UT, in order that you can see how the code works and use it in your own scripts, but a substantial number are effects which could easily be in a commercial, C-coded, plug-in set. Actually the included effects are the ones which will probably ship with Useful Things - they are still writing more, and there is at least a dozen 3rd. Party UT scripts not included here....
If you think UT scripts shouldn't be included then subtract 70, which gives:
Total filters available now:
=777
Total filters available for OS-X now:
= 467
If you include announced Betas:
= 564
Data like this will be available from AE itself, when it has full Effects Management! :-)
The situation is less clear with Photoshop filters which worked under AE.
There used to be at least another 200 which worked (enough to get by) in AE 5.0, although I suspect this number dropped substantially under AE 5.5, and I would guess that in AE 5.5 under OS-X the answer will currently be "none". If anyone is an early adopter of PS-7, then maybe they could try dropping the PS plugs in the AE filter folder and see if they work. I suspect not.
HTH
Dik
------------------------------------------------------------------------------------------
Message-ID: <009c01c1eb2d$73c8f510$303d03d2@KINESISADMIN.local>
From: "Kinesis-Eidolon" <kinesis@kinesis-eidolon.com>
Subject: Re: [AE] Plug-ins
Date: Wed, 24 Apr 2002 09:14:24 +0800
Ron wrote:
> Is there a one-stop
> set of reviews somewhere?
Not that I know of. I think it would be a very valid exercise, but (speaking from experience, since I've been keeping a database of available sets, initially aimed at allowing me to call and cross-reference available plug-ins into sensible Effects Management categories) it would be a monolithic one.
Consider the challenge:
- there are going to be about approaching a thousand filters which should be included in such a review
- to be able to review them fairly you have to have a fairly good handle on what they do, and there are a lot of filters around now which have close to 128 parameters and which do a whole stack of things depending on what parameters you set. Some filters also have listbox mode selections, so that they change behaviour quite a lot (each one of the Plug-in Galaxy filters - hot topic! - is actually several cheap and nasty filters crammed into one!)
- speed comparisons are extremely difficult since sophisticated filters can do a whole stack of stuff, and it depends what you are doing with them. Even relatively simple filters like blurs can't easily be compared (e.g. the ICE Gaussian and AE Gaussian blurs set at the same value don't look the same - so, to compare speed accurately you'd have to visually match them and then compare, since we don't know what they think settings like "3" actually refer to.
> I've heard that the Tinderbox filters are speedy.
Well that's not necessarily true. You can say that The Foundry's engineers generally write efficient code, but there are filters in the Tinder sets which are slow, because what they do requires a lot of processing. I suspect one area where ICE will still score is in things like Radial Blurs, which I guess will still be faster than most (haven't done comparative tests though.)
Generally, although you can say DE's code tends to be a bit slow, and The Foundry's is generally faster, you can't actually say "this set is all slow, this set is all fast". It has to be done on a filter by filter comparison.
> I know Shine is
> speedy
Case in point: Shine is a single filter "set" and you can thus make fairly good comparisons from it ;-)
> and looks great because I have a $7,000 set of useless things in > an old G4 450 (including Iced Lightburst). What sets are the slow dogs?
q.v. ^
> I got the OSX Delerium upgrade but it seems as slow on my dual 1G as it > did on my 450. Does anyone know if it's Altivec enhanced or uses dual > procs?
There are some slow filters in Delirium (compare Sketchist to other Sketch filters for example - although again look at bit closer - for instance The Foundry's sketch filter is substantially un-antialiased....)
On the other hand there are some which aren't - the DE Muligradient filter is quite snappy and bears comparison to others.
> Are the Boris plugs fast?
I would say, on average, yes. But for sure there are some which aren't. Again it depends on what they are doing - you would have to compare their particle systems against other particle systems which do similar particles
> Any fast particle systems out there?
Impulse Illusion ;-) - not an AE filter, but in Combustion 2
> Is the
> TInderbox blur faster than PB blur?
If you're talking the Gaussian blur (I thought all Adobe blurs were Standard ??)
then for sure the answer would be "yes" since I don't know of any Gaussian blur which is as slow as the Adobe one! T_Blur is exceptionally fast for larger area blurs, but there's trade offs in any algorithm, and for subpixel use - taking speed and "final look", then you may prefer the AE Gaussian sometimes - e.g. for subpixel use. Like I say these things aren't that simple to catelogue - and may even change with things like layer size (ICE suddenly loses it's speed over 768x576 for example.....)
HTH
Dik