Benoit Mandelbrot Fractal Art Contest 2009

[lkmitch]

From Damien Jones' posting to the Ultra Fractal mailing list:

*****
The Benoit Mandelbrot Fractal Art Contest 2009 is open and accepting submissions. This is a contest to select images for exhibition at ICM in 2010. The submissions period closes October 10, 2009.

For complete information, please visit the contest web site:

http://www.fractalartcontests.com/2009/
*****

Disclosure:  I am on the selection panel.

Kerry Mitchell

[cruelanimal]

Some of us in the fractal art community have raised issues about the manner in which the Benoit Mandelbrot Fractal Art Contest is run. While blogging at Orbit Trap, we have written extensively to detail why the protocols of this competition should be carefully examined with regard to professionalism, favoritism, and conflicts of interest.

This recent post might be especially noteworthy:

http://orbittrap.blogspot.com/2009/08/has-2009-benoit-mandelbrot-fractal-art.html

There are many more postings expressing concerns about the current and previous iterations of this particular contest.  Search the blog for “BMFAC.”  Here are just a few other examples:

http://orbittrap.blogspot.com/2007/11/price-of-professionalism.html

http://orbittrap.blogspot.com/2008/11/big-changes-proposed-for-benoit.html

http://orbittrap.blogspot.com/2007/09/questions-about-benoit-mandelbrot.html

A forum should be a place where all points of view can be freely expressed and explored.  I hope you will visit Orbit Trap, consider our lines of reasoning, and draw your own conclusions.  Thank you.

Terry Wright
Tim Hodkinson

Orbit Trap
http://orbittrap.blogspot.com

[David Makin]

Some of us in the fractal art community have raised issues about the manner in which the Benoit Mandelbrot Fractal Art Contest is run. While blogging at Orbit Trap, we have written extensively to detail why the protocols of this competition should be carefully examined with regard to professionalism, favoritism, and conflicts of interest.

This recent post might be especially noteworthy:

http://orbittrap.blogspot.com/2009/08/has-2009-benoit-mandelbrot-fractal-art.html

There are many more postings expressing concerns about the current and previous iterations of this particular contest.  Search the blog for “BMFAC.”  Here are just a few other examples:

http://orbittrap.blogspot.com/2007/11/price-of-professionalism.html

http://orbittrap.blogspot.com/2008/11/big-changes-proposed-for-benoit.html

http://orbittrap.blogspot.com/2007/09/questions-about-benoit-mandelbrot.html

A forum should be a place where all points of view can be freely expressed and explored.  I hope you will visit Orbit Trap, consider our lines of reasoning, and draw your own conclusions.  Thank you.

Terry Wright
Tim Hodkinson

Orbit Trap
http://orbittrap.blogspot.com

Been there, done that, still see no problem with The Contest

And yes I realise you'll probably just say "he uses UF and he was a winner twice so he would say that" but that's completely irrelevant as far as I'm concerned - I'd still be happy with the contest rules/organisation if I primarily used say Fractint or ChaosPro or FE and if I'd never won.

[cruelanimal]

Dave,

I don't know whether your being a previous two-time winner is relevant or not.  It was you who raised this point and not me.

However, questions about Ultra Fractal being privileged in the Benoit Mandelbrot Fractal Art Contest should be asked.  At least half of the 2009 judges have deep roots in the UF community.  By far, the majority of images exhibited in the first two competitions were made using UF.  The unnecessarily massive file sizes required to enter the contest seem deliberately designed to make sure that some software and fractals (like flames) will be unable to compete.

Therefore, I suspect that if you actually did use any of the other programs you mentioned, your odds of winning would be dramatically reduced.

There's an easy way to test the credibility of my suspicion. First, examine all of the images exhibited in the previous two contests.  Then, I'll make you a wager that there were more UF images shown in the first two exhibitions than images made with all of the following fractal programs combined: Fractint, Chaos Pro, Fractal Explorer, Xaos, Xenodream, all software by Stephen C. Ferguson, and all software by Terry W. Gintz.

And I could probably also throw Apophysis into the list and still win the bet.

Best,

Terry

[David Makin]

I don't know whether your being a previous two-time winner is relevant or not.  It was you who raised this point and not me.

Well if I hadn't and you didn't then I reckon someone else would have

However, questions about Ultra Fractal being privileged in the Benoit Mandelbrot Fractal Art Contest should be asked.  At least half of the 2009 judges have deep roots in the UF community.  By far, the majority of images exhibited in the first two competitions were made using UF.

I'd guess at least half the (serious) fractal artists are UF users
I think that the percentage of images chosen as winners that were UF based matches or is less than the percentage of entries that were UF based.

The unnecessarily massive file sizes required to enter the contest seem deliberately designed to make sure that some software and fractals (like flames) will be unable to compete.

No they're simply designed to get good quality 40" (or larger) prints. Myself I won't even contemplate less than 300ppi - I use 3840*2880 for my (10.66"*8") A4 prints.
As to the requirement for large prints, that's down to the fact that they're to be exhibited in a very large conference, not in someone's front room.

Therefore, I suspect that if you actually did use any of the other programs you mentioned, your odds of winning would be dramatically reduced.

I don't know about FE as I've never used it "in anger" but I believe large print quality renders are possible with both Fractint and ChaosPro.

There's an easy way to test the credibility of my suspicion. First, examine all of the images exhibited in the previous two contests.  Then, I'll make you a wager that there were more UF images shown in the first two exhibitions than images made with all of the following fractal programs combined: Fractint, Chaos Pro, Fractal Explorer, Xaos, Xenodream, all software by Stephen C. Ferguson, and all software by Terry W. Gintz.
And I could probably also throw Apophysis into the list and still win the bet.

All entered images are on the contest sites. You are correct there are probably more UF *entries* than all the others combined - that's not something that the contest itself has control over - how many images did you and Tim submit ?

[lycium]

i knew the drama queens would be back soon after the announcements were made...

+1 vote that you trolls go take your stinking controversy elsewhere 

[cruelanimal]

Dave,

If "at least half the (serious) fractal artists are UF users,"  then why have the other half of the (serious) non-UF fractal artists had no comparable representation in past exhibitions.  It looks to me like the contest has tried to suggest that nearly all (serious) fractal artists use Ultra Fractal -- almost by default.  There are more UF entries in the contest because it, being filled with past UF judges and artists, gives the impression of being heavily influenced if not outright favoring UF.

The exhibition prints do not have to be made to be seen from space.  Conference or not, this is an art contest presumably being shown in a gallery setting and not in a boardroom.  Good fine art comes in all sizes, and you and I both know that archival fine art prints can be made with well under the contest's huge entry dimensions.  The size restrictions seem to be in place to deliberately make entering difficult for some artists.  This is supposed to be a contest for all fractal art, and not just one that represents work that can "measure up" to a plasma TV.

The contest may not have control over what it gets, but it can select judges and make rules to see that it gets what it wants.

Best,

Drama Queen

[David Makin]

> Conference or not, this is an art contest presumably being shown in a gallery setting and not in a boardroom.  Good fine art comes in all sizes, and you and I both know that archival fine art prints can be made with well under the contest's huge entry dimensions.

In terms of inches I would agree that archival fine art prints do not need to be so large but I suspect that in this case the "gallery" is a large conference hall not an intimate art gallery.

[cruelanimal]

Dave,

But in the 2007 competition, which was not tied to a conference and was held in a more intimate gallery setting, the same mammoth sizes were a requirement for entering the contest.  So, the size restrictions are in place for another reason: to thin the herd and to favor particular programs over others.

Best,

Terry 

[David Makin]

Errm it was in a gallery but I wouldn't call it an intimate one - they were large rooms suitable for large images.

[cruelanimal]

Dave,

I included a photograph of the 2007 BMFAC exhibition in this Orbit Trap post:

http://orbittrap.blogspot.com/2008/08/review-of-week-2007-benoit-mandelbrot.html

This is the only photograph of the exhibition I could find on the Net.  Oddly, even the 2007 contest site has no photographs of the exhibition.  In fact, it has absolutely no account of the exhibition at all.

Note that the three works are hung closely together in what clearly seems to me to be an intimate arrangement.

Again, I maintain the gigantic size stipulations are in place to deliberately privilege artists who use programs like Ultra Fractal but lock out artists who use other programs -- especially Apophysis (since flames are difficult to render at such huge sizes), artists who heavily post-process, and artists who simply don't have the resources to afford the needed hardware firepower to scale to barn door size.

Best,

Terry 

[David Makin]

Dave,
<snip>

Again, I maintain the gigantic size stipulations are in place to deliberately privilege artists who use programs like Ultra Fractal but lock out artists who use other programs -- especially Apophysis (since flames are difficult to render at such huge sizes), artists who heavily post-process, and artists who simply don't have the resources to afford the needed hardware firepower to scale to barn door size.

Best,

Terry 

"Flames are difficult to render at such sizes" - actually that's balderdash, I wrote MMFrac v2.4 about 8 years ago and that allows an 8000*6000 RAM buffer (yes RAM not VRAM) on a PC with 512MB of RAM (in 32-bit DOS) so it would be fairly simple to add render to large buffer to Apo especially since few folks nowadays have less than 2GB RAM.
"artists who heavily post-process," - I don't do it often but have successfully post-processed images at 8000*6000 in PSP (OK sorry - that's commercial) and in Paint.net (free).
"don't have the resources" - if you're serious about fractal art it's easy enough to get the resources, I earn so little that I haven't even had to pay tax (except once about 10 years ago) and I have a Desktop PC with 2GB Ram and a laptop with 3GB RAM and a copy of UF, Xenodream and about 20 free fractal programs - I don't see any other resources being required except maybe Adobe Photoshop (which I don't have) or Paint Shop Pro (of which I have version 6). Just to add that I'm fairly sure that (legal) up to date copies of Adobe Photoshop or Corel's PSP are both more expensive than either UF or XD and most serious artists of any digital type own at least one of those (or a similarly priced equivalent).

[cruelanimal]

Dave,

So my claim that it’s difficult to render flames at large sizes is “balderdash.”  It’s somewhat ironic, then, that the following statement appears to contradict you:

“Some images may be selected for printing at even larger size (12000 pixels in the largest dimension) so entrants would do well to be aware of the size requirements. This is particularly important for certain types of fractals (e.g. flames) which are difficult to render at large sizes.”

Do you recognize the source?  It comes from the “Entries” section of the Rules Page for the 2009 Benoit Mandelbrot Fractal Art Contest web site.  You can refresh your memory here:

http://www.fractalartcontests.com/2009/rules.php

Moreover, contest co-director Damien M. Jones referred to this very topic when making forum announcements about the 2007 BMFAC and wrote:

“Because our goal is to select artwork for exhibition, you need to be able to produce a high-resolution version of your image. Usually, this means at least 8,000 pixels on the shorter side. If you cannot produce a large version of the image, you should not enter that image. (This tends to be a problem with Apophysis images especially; you may enter them, but keep in mind that very large Apophysis renders take an extremely long time. Start now.)”

It sounds like Jones even finds the Apo render difficulties joke-worthy.  By the way, a copy of the 2007 announcement can still be seen here:

http://news.deviantart.com/article/29528/

I somehow doubt that you and Jones (who may have also written the rules text) can both be right.

As for post-processing, frankly, from seeing your work, it looks to me like you do nearly all your processing inside fractal programs.  In fact, I would bet you all the tea on Teavana.com whatever post-processing you’ve done is infinitesimal compared to the amount I regularly do.  Have you ever really done any heavy post-processing at 12000 pixels using commercial filters in Photoshop CS 2 or 3 without at least duo processors and four gigs of RAM?  Try it because I predict you won’t find it a zippy experience.

Finally, I suppose it’s easy enough to get the resources if you’re serious about fractal art.  What’s difficult, for some fractal artists anyway, is to get (or own) the resources to render fractal art to the dimensions demanded by this particular contest.  That, I believe, is the point at hand.  You’ve really only shown so far that you have adequate resources to meet the size restrictions and can therefore enter the 2009 BMFAC.  Maybe I’m just missing your point.  Are you are arguing that the only fractal artists who are “serious” enough to enter the contest must at least have a machine as powerful as yours and like you possess copies of both Xenodream and Ultra Fractal – programs that just coincidentally happen to scale images to the large sizes insisted upon by the competition?  If so, we’ve gone full circle, so I’ll say again: The contest’s size restrictions look designed to lock out some hardware/software-deficient (and thus obviously un-serious) fractal artists.

And since you seem to have taken it upon yourself here to be the apparent de facto spokesperson for the competition, do you have any thoughts about the decision to have two authors of commercial fractal software serve as judges in this year’s BMFAC?  Don’t they stand to benefit monetarily if images made with their software are selected as winners?  And can you further explain to me why having them serve on the selection panel is a good thing – rather than a self-evident conflict of interest?

Best,

Terry

[lycium]

Terry, you should find concern in the fact that people much younger than you regard you as childish.

Yes, it's true that Apophysis is a slow 32bit Delphi program, which in Windows can only access 3GB* of memory and therefore is quite inefficient at rendering very large images via the multi-pass process. That's not anyone's fault, it's just a consequence of the (unmodified) algorithm's behaviour - certainly not some evil doing by Damien to reduce the number of Apophysis entries.

Maybe instead of posting your elaborate conspiracy theories year after year, you could do something meaningful to address the inequities you perceive - like making good artwork and not submitting it to the contest. Stop being so bitter and wasting your energy on trying to belittle the competition; rather use it to improve your skills and standing, or organise a better venue for "equal and fair" fractal art competitions. Your whining certainly isn't doing you any favours.

Best,


Thomas

* Actually Michael Faber recently chided me for my remark that 32bit Windows programs have a 2GB process limit; Apophysis is compiled as Large Address Aware so they can access 3GB, whoop dee doo.

[David Makin]

Dave,

So my claim that it’s difficult to render flames at large sizes is “balderdash.”  It’s somewhat ironic, then, that the following statement appears to contradict you:

“Some images may be selected for printing at even larger size (12000 pixels in the largest dimension) so entrants would do well to be aware of the size requirements. This is particularly important for certain types of fractals (e.g. flames) which are difficult to render at large sizes.”

Do you recognize the source?  It comes from the “Entries” section of the Rules Page for the 2009 Benoit Mandelbrot Fractal Art Contest web site.  You can refresh your memory here:

http://www.fractalartcontests.com/2009/rules.php

Moreover, contest co-director Damien M. Jones referred to this very topic when making forum announcements about the 2007 BMFAC and wrote:

“Because our goal is to select artwork for exhibition, you need to be able to produce a high-resolution version of your image. Usually, this means at least 8,000 pixels on the shorter side. If you cannot produce a large version of the image, you should not enter that image. (This tends to be a problem with Apophysis images especially; you may enter them, but keep in mind that very large Apophysis renders take an extremely long time. Start now.)”

It sounds like Jones even finds the Apo render difficulties joke-worthy.  By the way, a copy of the 2007 announcement can still be seen here:

http://news.deviantart.com/article/29528/

I somehow doubt that you and Jones (who may have also written the rules text) can both be right.

As for post-processing, frankly, from seeing your work, it looks to me like you do nearly all your processing inside fractal programs.  In fact, I would bet you all the tea on Teavana.com whatever post-processing you’ve done is infinitesimal compared to the amount I regularly do.  Have you ever really done any heavy post-processing at 12000 pixels using commercial filters in Photoshop CS 2 or 3 without at least duo processors and four gigs of RAM?  Try it because I predict you won’t find it a zippy experience.

Finally, I suppose it’s easy enough to get the resources if you’re serious about fractal art.  What’s difficult, for some fractal artists anyway, is to get (or own) the resources to render fractal art to the dimensions demanded by this particular contest.  That, I believe, is the point at hand.  You’ve really only shown so far that you have adequate resources to meet the size restrictions and can therefore enter the 2009 BMFAC.  Maybe I’m just missing your point.  Are you are arguing that the only fractal artists who are “serious” enough to enter the contest must at least have a machine as powerful as yours and like you possess copies of both Xenodream and Ultra Fractal – programs that just coincidentally happen to scale images to the large sizes insisted upon by the competition?  If so, we’ve gone full circle, so I’ll say again: The contest’s size restrictions look designed to lock out some hardware/software-deficient (and thus obviously un-serious) fractal artists.

And since you seem to have taken it upon yourself here to be the apparent de facto spokesperson for the competition, do you have any thoughts about the decision to have two authors of commercial fractal software serve as judges in this year’s BMFAC?  Don’t they stand to benefit monetarily if images made with their software are selected as winners?  And can you further explain to me why having them serve on the selection panel is a good thing – rather than a self-evident conflict of interest?

Best,

Terry

Well for one thing I disagree with Damien in the sense that rendering large flames is difficult - it's not dificult at all, it just can be a little slow - but then again so are many fractals I render, my print release version of "Aquamarine" took a solid month of computer time (and that's a normal escape-time fractal).
Here on the subject of time, given your comment about image processing in PSP or whatever, it simply sounds like you need to be a lot more patient - I'd expect each change I make to a large image (>=8000*6000) to take up to an hour, sometimes changes I make to my escape-time fractals at 640*480 take half an hour to render (I mean while I'm working on them not rendering to disk).
I confess I haven't tried it in Apo but all it needs is a suggestion to the current developers of Apo to optimise large image rendering - or better still code it up yourself (use integers instead of floats for a gradient colour value and #hits value-just 16-bits for each would still produce good images).
Rendering flames in UF is difficult but this is because UF was not really designed with flame fractals in mind, it was designed for escape-time fractals - hence my escape-time formulas for rendering 2D and 3D IFS.

Your comment about "big and powerful rigs" certainly does not apply to me - the machine I do most of my work on is a 3GHz P4HT and given that I'm a fractal artist I'm slightly ashamed to say so. I should point out that entry-level PC's now have the same amount of memory that I do (2GB) and are around 4* faster (being core2Duo). In fact I'm positive you could quite easily get a second-hand set-up for under £150 that would vastly out-perform what I'm using.

On the subject of the contest and judges - I believe that having such a contest was not exactly essential as far as the Mathematicians Congress was concerned, the images chosen for exhibition could have been by invitiation only - perhaps you would have preferred that (i.e. no contest at all) ?