Talk:Random dot stereogram

The Random Dot Stereogram is a concept much more important than Autostereogram. It has to do with science, not just entertainment. The Autostereogram page is too long to explain a simple concept. And it has too many images only for entertainment. It is ok for an entertainment page. That is why I think the Random Dot Stereogram should be kept in its own page. For the memory of Dr. Bela Julesz. —The preceding unsigned comment was added by Otuyama (talk • contribs).


 * A random dot stereogram is a kind of autostereogram, so it doesn't make sense to say it's "much more important". What do you mean by "an entertainment page"? I think Autostereogram explains the concept much better than this article. —Keenan Pepper 04:01, 27 May 2006 (UTC)

I think this is more important because it is a scientific proof. It shows how retina and brain works. It needs to be random dots to avoid any clue (shape from shading, sobreposition, etc). - --Otuyama 18:38, 27 May 2006 (UTC)


 * What do you think is more important than what? Be specific. We're talking about whether these concepts would better presented as one article or two. I don't understand how your comments are pertinent to that question. Also, what kind of word is sobreposition? It appears in this article, so I doubt it's a typo, but I've never seen it before and it's not in the OED. Is it from another language or something? —Keenan Pepper 20:45, 27 May 2006 (UTC)


 * Oh, and let me apologize for my error: a two-image random dot stereogram as described in this article is not a kind of autostereogram, because autosterograms are single images. That's a better argument for not merging the articles than anything you've said. —Keenan Pepper 20:51, 27 May 2006 (UTC)

Autosterogram is important because it made the concept of random dot stereogram widely public, for non scientific public. It happened because it is a very interesting illusion. But it is not a scientific proof. The random dot stereogram is a scientific proof. It has an axioma: random dots do not convey any kind of three dimensional information (shape from shading, contours, etc). If you present only one image, you cant say what is the hidden image (it is only random dots, see axiom). So neurons at retina are not responsible for the stereopsis. But when you see both images, you can say what is the hidden image - stereopsis occur somewhere in the brain. You can say more: somewhere in the brain that receives information from both eyes. That is the proof of Dr. Bela Julesz, with my words. Sorry, I could't find the spell checker in this editor. I apologize for the mistake.


 * I'm very confused. Are you talking about the concepts or the articles? What exactly do you think an autostereogram is? —Keenan Pepper 23:18, 28 May 2006 (UTC)

I talking about the concepts and what it results in the articles. I believe an article is about the concept it tries to explain. The diference between the random dot stereogram and an autostereogram is that the first has two images and random dots, and the second is a continous image of patterns (maybe random dots). It is not just a matter of having one ou two images or having random dots or coloured patterns. The random dot stereogram proof requires two images and random dots. --Otuyama 07:11, 30 May 2006 (UTC)

Strictly speaking, random-dot stereogram refers to two images, while autostereogram refers to a single image with repeated patterns which employ the same principles behind RDS. The problem is that nowadays people often refer to autostereograms as stereograms. This is why I tried very hard to get names right in the autostereogram article. See the Terminology section for details. I'll change the 'Random dot stereogram' entry in the Terminology section to link to this article. Fred Hsu 02:23, 1 June 2006 (UTC)

I just edited the lead section to clarify the difference between RDS and SIRDS. I just added a new discussion section where people to discuss how various types of autostereograms can be classified. Fred Hsu 02:47, 1 June 2006 (UTC)

Better cut out please

 * I wish the contributor of the stereograms hadn't just made matching blobs. A little stick man or a heart might be better. Anthony717 (talk) 00:54, 23 March 2008 (UTC)


 * Do check out Autostereogram. Hopefully you'll find better examples there. Fred Hsu (talk) 01:54, 24 March 2008 (UTC)

A great link
Single image random dot stereogram (SIRDS) explained: http://www.inf.ufsc.br/~otuyama/eng/stereogram/basic/index.html

In short: use PhotoShop; make selection of the object(s) to be viewed stereoscopically; move horizontally the selected portion of the previous pattern strip to create the next one - each moved the same amount to the right/left (repeated; the first one is of course the original pattern strip). At the end you get an array of pattern strips side by side.