hatena
<body>
*1246276684*Tried Yodobashi Camera's e-Photo Net service.
f:id:nishiohirokazu:20090629210138j:image
The edges are cut off by about 1mm. Well, it is not so bad for the purpose of experimenting before printing more expensive prints, or for printing CG images to show to others in situations where there is no PC available.
*1246279159*[svg]recursive KOIKE
<img src="http://gyazo.com/c057e46da2f1ab9cedf089d97890e58f.png">
SVG memo
>||
<defs>
<symbol id="Koike">
...
||<
Define Koike with
>||
<use xlink:href="#Koike"/>
||<
Koike is used in
Easy. The reason it doesn't look very regular is probably due to the fact that the two SVGs were originally created independently. I think it would be better to first decide on the function to be used for mapping, then create a converted image of a square with a grid, etc., and place the picture on the grid. It would be just like a texture painter trying to apply a standard texture first.
</body>
<comments>
<comment>
<username>bgnori</username>
<body>Simple question: What do you think about the resolution of the output and the noise caused by lossy compression? </body>
<timestamp>1246280423</timestamp>
</comment>
<comment>
<username>nishiohirokazu</username>
<body>I'm not sure what you mean by "what do you think" in your question, but the resolution is 360 DPI and the JPEG compression is the lowest compression ratio with the least noise. </body>
<timestamp>1246281222</timestamp>
</comment>
<comment>
<username>n_shuyo</username>
<body>A4 size is ¥195 at vivipri. If you need to trim the image, you can specify the area when you order. </body> <body
<timestamp>1246281480</timestamp>
</comment>
<comment>
<username>bgnori</username>
<BODY>I suspect that jpeg is not good for what you want to output because the resolution is high and it is like a line drawing, and I am guessing that there is a part of you that intends to produce it in detail as a work of art. I wanted to know how that particular aspect is pursued. </body>
<timestamp>1246288119</timestamp>
</comment>
<comment>
<username>nishiohirokazu</username>
<BODY>I'll try it next time with vivipri. <BR><BR><BR>I'm not sure about the difficulty with high resolution. It is often said that low-resolution images have noise in blocks of 8 pixels, but this one has a resolution of about 16 pixels per millimeter, which means that the block boundaries are in 0.5-millimeter increments. <br>That is why I am not dissatisfied with the image quality, but I am much more dissatisfied with the fact that the edges have been cut off by 1mm, since the left foot of the Virtual Star was designed to meet the bottom left corner, and the line drawing style is also off, since the bottom line converges to 0 when it should be observable. I am rather disappointed that the line drawing style has been cut off, even though it should be possible to observe the convergence of the bottom line to 0. </body>
<timestamp>1246298587</timestamp>
</comment>
<comment>
<username>shou</username>
<body>Perhaps what bgnori is trying to say is that higher resolution = higher spatial frequency. <br>I think what bgnori is trying to say is that high resolution = high spatial frequency. <br>I think he is trying to say that with a method like jpeg, which quantizes in the frequency domain, the spatial frequency makes it difficult to reproduce the image. </body>
<timestamp>1246344421</timestamp>
</comment>
<comment>
<username>bgnori</username>
<body>I'm afraid I'm not asking the question with much understanding, but<br><br>1) The thing you want to draw is something that can be enlarged as much as possible in the first place, and you are talking about the intention of the creator to draw it to what level of detail (reproduce to high frequency) when people see it<br>2) I'm wondering if you can't achieve the intention of 1) by outputting the drawn thing as a jpeg file? br>2) The discussion of whether the intention of 1) cannot be achieved if the rendered image is output as a jpeg file. <BR><BR>I understood that for this A4 output, 1) is not really an issue, and 2) is that they think 16 pixel/mm and a block pitch of 0.5mm is sufficient. <BR><BR>I personally fantasize about what I might be able to see if I output at 1m x 2m. <BR><BR>I heard that professionals use tiff. <BR><BR>I heard that professionals use tiff. I'm not sure I can keep up. <br>http://www.genkosha.com/eizo/serial/boco/boco6.html</body>
<timestamp>1246363673</timestamp>
</comment>
<comment>
<username>nishiohirokazu</username>
<body>I also submit my finished prints in TIFF. There is no reason to go to the trouble of choosing a format that deteriorates. <BR><BR><BR>The reason I send my drafts as JPEGs to a photo print shop is because it's cheaper that way. <br><br><br>As for 1), the line drawing style is not a fractal structure, so I don't think it would be interesting to enlarge it any further. Currently, 10 recursions and A4 is too fine for the human eye to perceive as a "dot" instead of a "circle," and it doesn't look beautiful. Each recursion is 0.618 times deeper, so it would be possible to recurse more than 20 times in a gymnasium size. At that size, it cannot be seen from the front, which may have some interesting effects... I would like to try to make it on the whole wall like "No.3" by a "contractor who tried to make it" (http://blog.tap2.jp/), if there is a place for it. Cost and location are the main issues. The reason why my works are basically A3 is that if they are too big, I don't have a place to store them, and I can't transport them by myself, so the postage cost is high. The cost is not infinite. </body> </P> </P> <BLOCKQUOTE
<timestamp>1246366296</timestamp>
</comment>
</comments>
Hatena Diary 2009-06-29
This page is auto-translated from /nishio/Hatena2009-06-29 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I’m very happy to spread my thought to non-Japanese readers.