PercepTools a hoax?

Started May 28, 2009 | Discussions
Roland Karlsson Forum Pro • Posts: 28,726
PercepTools a hoax?

Hi,

On the front news right now is PercepTools. A press release proudly presents it and describes it as a simulation the perception of the human brain. And there is even a demonstration how it distinguishes between reflection and illumination and then only affects the latter. That demo is a piece of graphics art, not a photo.

There are also three examples - all B&W. Looking at the examples - I think it is a bluff. OK - the tool probably works just fine, maybe making some tone mapping. But to me it looks rather obvious that it is just that - tone mapping to improve contrast.

The leaf image is just lighter and has really no illumination map. So - it is bogus.

The egg image does not have any reflection map - so its bogus.

The forest image looks convincing. You can feel that the illumination contrast increases. Is this it? Does it really work. Nah .... I dont think so. There are really only two reflection gray tones - the trees and the rest. If the tool really can just emphasis the illumination map - then trees near to brighter lit areas should also get lighter. But they dont - they get darker. Just as all gray tones are affected by the same tone mapping.

So - maybe it is a cool and good tool, but the description on what it does is probably bogus.

-- hide signature --
 Roland Karlsson's gear list:Roland Karlsson's gear list
Sony RX100 III Sigma DP3 Merrill Sigma dp2 Quattro Pentax K-3 Pentax K-1 +14 more
ikillrocknroll Regular Member • Posts: 224
Re: PercepTools a hoax? Yes

I feel the same as you, and feel that I could easily go from the before to the after in photoshop without using that plugin.

Matthew Miller
Matthew Miller Senior Member • Posts: 1,311
paid advertisement?

While I understand that releases of major software packages are generally newsworthy, this one struck me as a bit odd. There's thousands of photoshop plugins available, and they don't all show up as News. Is it a paid advertisement?

-- hide signature --

Matthew Miller « http://mattdm.org/ »

 Matthew Miller's gear list:Matthew Miller's gear list
Pentax K-5 II Fujifilm X-T10 Pentax smc DA 40mm F2.8 Limited Pentax smc DA 70mm F2.4 AL Limited Fujifilm XF 23mm F1.4 R +5 more
Robert Deutsch Forum Pro • Posts: 10,227
Try this in photoshop

1. Open the "before" leaf image.
2. Go to Curves and click on Auto.

The result looks very similar to the "after" image. That image has a bit more contrast still, which I suspect that could be matched with a bit of tweaking.

There may be something more subtle that becomes apparent on closer examination, but so far I'm not convinced.

Bob

 Robert Deutsch's gear list:Robert Deutsch's gear list
Olympus E-M5 II
Photato
Photato Senior Member • Posts: 2,704
The 'Before' samples are underexposed....

If you check the 'before' samples, you'll see that they are underexposed.
The forrest for instance is like -1.5 EV !
So no FAIR comparison.

 Photato's gear list:Photato's gear list
Panasonic LX100 Canon EOS M Canon EOS M6 Canon EOS M6 II Canon EF-M 11-22mm f/4-5.6 IS STM +12 more
Andrew Westlake Senior Member • Posts: 2,928
Not an advertisement

Matthew Miller wrote:

While I understand that releases of major software packages are
generally newsworthy, this one struck me as a bit odd. There's
thousands of photoshop plugins available, and they don't all show up
as News. Is it a paid advertisement?

Absolutely not. You can tell, by the way, by the fact that it's presented as a news story, and not in one of the ad slots.

It's simply a plug-in for which we received a press release, and considered sufficiently interesting to mention. If you're not convinced, there's a simple answer - don't buy it.

-- hide signature --

Andy Westlake
dpreview.com/lensreviews

 Andrew Westlake's gear list:Andrew Westlake's gear list
Fujifilm FinePix F30 Zoom Olympus XZ-2 iHS Olympus Stylus Verve S Panasonic Lumix DMC-GF1 Olympus OM-D E-M5 +11 more
mousse Regular Member • Posts: 310
Re: PercepTools a hoax?

I tried it. Am not impressed.Nik has a gradual neutral density filter that can achieve

about the same and its fast. This PercepTool is very very slow, at least on my Mac.

Joe Marney
Joe Marney Regular Member • Posts: 178
Re: PercepTools a hoax?

Sounded cool so i tried it out.

What a waste of time. Just use curves and get the same result....

There are plenty of decent plug-ins out there--why promote this???

john Senior Member • Posts: 1,620
Re: PercepTools a hoax?

the difference i can only see is increased contrast, you can add 10% contrast in photoshop alone without any plugin.

Chris59 Forum Pro • Posts: 15,437
Re: PercepTools a hoax? No! Worse, it's misleading advertising.

No computer program can "know" what the illumination and reflectance of any object or any series of different objects is without knowing exactly what kind of lighting was used, in what room/studio and exactly what the objects are.

As for then taking that information and replicating the brain's way of perceiving the scene, well, the less I say about that, the better!

 Chris59's gear list:Chris59's gear list
Samsung NX1 Samsung 16-50mm F2.0-2.8
Robert Deutsch Forum Pro • Posts: 10,227
Give DPR a break

Joe Marney wrote:

Sounded cool so i tried it out.

What a waste of time. Just use curves and get the same result....

There are plenty of decent plug-ins out there--why promote this???

I'm not impressed, either, but I don't see DPR as promoting this. They're merely reporting on a plugin that comes from a legitimate source and that sounds as though it might be interesting. It's up to us to evaluate it.

Bob

 Robert Deutsch's gear list:Robert Deutsch's gear list
Olympus E-M5 II
todd wms Regular Member • Posts: 140
Didn't do it for me

It may be enough processing for some, but to my eye and my own processing techniques, this tool (?) doesn't go far, at all. Certainly the results it achieves appears to me to be fairly rudimentary stuff. It's not like he's bending light and taking the image to the 4th dimension, seriously now...

Got plenty of ways I'd rather spend my $90. Percep Tool doesn't appear to be worthy of anywhere, not ANYWHERE NEAR, that amount of money, and his rationalizations are quite self congratulatory, pumped up. It's like he's already speaking to criticism.

But that's just me..........

-- hide signature --

Todd Williams

PicOne
PicOne Veteran Member • Posts: 6,932
PercepTools = AutoLevels? nt
-- hide signature --

'Everything in photography boils down to what's sharp and what's fuzzy.'
-Gaylord Herron

tko Forum Pro • Posts: 13,140
how many other plug in's has DPR covered?

How many plugins are there on the market?

How many good ones are there that do something useful?

How many has DPR headlined?

Is this plugin useful?

Seems like the answers are a lot, some, one, and no.

The people are DPR are pretty smart. I'm sure they know about Neatimage, Photomatix, FocusMagic, and a host of others. If you're a good reporter you dig out the worthwhile news, you don't just post the garbage people send you w/o any critical thought.

gpedder Regular Member • Posts: 118
Re: how many other plug in's has DPR covered?

The explanation given as to the inner working of the plugin is indeed bull. There's no way to split an image like that, the human brain doesn't work like that, it's absolute codswallop. Surely the dpreview editors know this?
--
Geoff pedder

todd wms Regular Member • Posts: 140
What confuses me Andy...

... is why you choose this plugin to report?

I bet it's safe to say that every responder in this thread took the time to download and install; and then in figuratively 2 minutes, this PercepTool was debunked.

I'd say it appears from the comments that all of us seem to concur that the effort of any curiosity at all does not find this plugin "sufficiently interesting". And when contrasting that discovery to the many fine announcements about anything digitally photographic that DPR wonderfully reports about to us (and thank you very very much), we just come up wondering why this one..

The plugin smelt like a plug. Sorry.
--
Todd Williams

Robert Deutsch Forum Pro • Posts: 10,227
Re: how many other plug in's has DPR covered?

tko wrote:

How many plugins are there on the market?

How many good ones are there that do something useful?

How many has DPR headlined?

Is this plugin useful?

Seems like the answers are a lot, some, one, and no.

The people are DPR are pretty smart. I'm sure they know about
Neatimage, Photomatix, FocusMagic, and a host of others. If you're a
good reporter you dig out the worthwhile news, you don't just post
the garbage people send you w/o any critical thought.

I don't really want the role of apologist for DPR, but the plugins you refer to have been around for quite some time, whereas PercepTool is new, and the section of the web site where the description appears is "Latest News." I don't see how DPR can be expected to evaluate the usefulness of this software in what it clearly a news item. They just reported that it's available, and presented information based on the press release, without any evaluative comment. I never got the impression that DPR was endorsing PercepTool.

Bob

 Robert Deutsch's gear list:Robert Deutsch's gear list
Olympus E-M5 II
ALbino Regular Member • Posts: 150
Original Egg Image, Contrast = 65%. Virtually identical

I chopped the Egg image in half, left one side as the Before and the other as the After. Then I changed the contrast on the Before to 65%. You can just make out the line between the two pictures.

Here's their original comparisons:

And here is Before + 65% Contrast in comparison to their After:

If there is a difference it's point escapes me.

G10Rebel
G10Rebel Contributing Member • Posts: 788
Quite harsh, but...

... probably Topaz Adjust can do it too, perhaps better, but surely cheaper
--
http://hdr.strivearth.com | canon fodder

PhotoArtKC Senior Member • Posts: 1,585
Re: PercepTools a hoax?

I downloaded the demo and gave it a try on a few images.

Since I do color correction for professional orders at a major US lab, I can pretty much get the color, density and contrast pretty much where it should be on my own. But hey, why not try right?

So I gave it a test on a few images. A few already corrected while a few were slightly dark and flat still. Overall, all it did was boost highlight contrast often blowing out details in whites on properly corrected images, and slightly lighten the image. No magic that can't be done for free with a VERY simple curve in Photoshop in far less time than it took the plug-in to process my images.

On top of that, it took about 1 minute per image to process mid size (about 7mp cropped) 8 and 16bit TIFF files. That on a dual processor Mac Pro with plenty of power to spare.

I do feel a little sorry for the flak DPR is taking over the "news". I suspect someone should/will be testing software and plug-in's for validity to claims in the future. hint hint Needless to say, the software is clearly snake oil and not worth 89 cents much less $89.

-- hide signature --

There are no rules for good photographs, there are only good photographs. - Ansel Adams

My photography blog, Photoshop tips, color tips and more: http://modifiedphoto.wordpress.com

Keyboard shortcuts:
FForum MMy threads