Could deletions be more specific?

Tagged under Burst Angel

page 1 of 1 9 total items

Astara

Astara

Scanner, Artistic

Cancel

Athenae,

The scan you have submitted (Jo, Avenging Angel) was removed for one or more of the following reasons:

- The image was too small. Only scans with a clear subject of 1,000,000 pixels or more are accepted.
- The scan contains excessive artifacts. This includes, but is not limited to: page folds, dust, or text on the image.
- The image was enlarged. The majority of anime images found on the internet are not sufficient quality to be submitted at Minitokyo, please do not attempt to enlarge these images for submission.

---------------

The above is pretty vague.

It was self-scanned. It was approved as being a higher resolution and better quality image -- and replaced the original scan. And then it gets deleted for some vague reason by someone else?

Note my scan was It was 3247x4469, at 400 dpi. This corresponds to the original at 8.11 x 11.17", which is the size of the artbook, so it obviously wasn't enlarged.

Could someone please clarify what the problem is? I spent a fair amount of time getting this scan adjusted to the quality you see. I didn't just slap it on the glass and upload it, but I tried multiple settings, exposures, reprofiled and calibrated my monitor's color, and compared the calibrated result to the original to ensure an accurate result (I'm not always so careful, but some scans need extra attention to get a good result).

Below (if this works) are the two images, mine to the left, the one in the gallery to the right, both shown at the same size.

http://img46.imageshack.us/img46/4164/avengingangelvsoldersca.jpg

If that doesn't display right, here's a link.

Notice in the one on the right that Jo's skin, is off-color -- way off. This isn't true in the one on the left (nor in the original art). Look at the faces -- the one on the left is noticeably sharper than the one on the right. So why was mine deleted? The older scan is off color, was overscanned by 200% (i.e. enlarged by 200%), shows more moire patterning at high zoom and is more blurry. So could someone be more specific about why mine was worse than the previous one and qualified for deletion?

In general, if the canned response doesn't clearly answer the question, it would be nice to to elaborate on the reasons for deletion.

Thanks,

fireflywishes

Retired Moderator, Linguistics

fireflywishes

Calgon, take me away~!

Cancel

Read past the automated message part... the Mod who deleted it should have posted their own comments at the end.

Signature Image

Astara

Astara

Scanner, Artistic

Cancel

The above was the entire text of the note.

If they'd posted more, I wouldn't be writing... I have seen more on other notes and it answered my questions. This one was very perplexing -- I think they were looking at the old scan or something and got confused...

The one on the right looks like it has had the color deliberately adjusted -- there is some solarization on the arm -- indicating an out-of-range color adjustment. Maybe they had an odd color profile set on their system that adjusted to be more intense, or maybe they worked on it with the adobeRGB color profile -- which can intensify color, I learned recently....

A*a

merged: 01-30-2010 ~ 01:52am

Um...since no one has actually responded as to 'why?' my original upload was deleted, (consistent with there being no details given as to why it was deleted, is it possible there was a system error or that someone thought there was a problem with my scan and now realizes they made a mistake and doesn't want to answer?

Should I reupload my copy...or can it be restore?

kayurachan

Retired Moderator

kayurachan

Mado&Fen are my Gods.

Cancel

Hi Athenae, sorry for a late reply.
I deleted your scan since it was reported for being double. I compared the two scans in Photoshop and your is just slightly bigger than Saikusa's and a lot blurred and overfiltered. Looks a bit enlarged too. You might want to re-scan it and post it again?
And please give us some days before opening new threads on your deletions, ok?

Signature
	Image
!Death to J-Rock scans! >.<
What is Kayurachan doing? Just see it on my userpage :3
Sheq says: XD one sec, eating an egg XP

Astara

Astara

Scanner, Artistic

Cancel

It's more blurred when compared at the 33% greater size. If you compare them at the same size, as I did above, mine is sharper. The purpose of submitting them at a slight larger size, was because *before* (on a different scan), I had another that was sharper (at same size view) Mine was 300 dpi of an 8.2x11.5 print compared to the older db-version that was 600dpi of a 4.8x7.2" print. The original prints both appeared to have been printed at 300dpi. So scanning at >300dpi resulted in increased pixelization that was present in the 'in-db' version.

My preference was for quality at 300dpi, vs. an overscan at 400 or 600, but since that one got deleted, I thought it was board policy to only look at dimensions. So on this one I used a 400 dpi scan of the print giving about 20% greater dimensions, and tried to slightly smooth out the resulting pixelization in PS -- resulting in (doing side by side comparisons) an image that was less blurred, less pixelated, and not color-damaged as the one in the db is. But it still got deleted.

So how do I communicate the quality level when of a scan compared to a damaged-db copy when they are not easily evaluated under equals conditions that are **suitable** for use in a wall (presuming that is the reason for the 'scan' DB).

I don't like submitting pixelated scans have have given up on submitting some pictures if I couldn't exceed the dimensions of the one present w/o pixelization. However, on some the color damage was so bad I downloaded the db copy to see if it was just a matter of a bad color profile or what. (Color profile mismatches can make picture colors look way off, but I'll leave that at that, as that's a whole book by itself!)

What I found was it's resolution wasn't as high as its numbers indicated -- but that it was been scanned in at 2X the source-image's resolution @ 600 dpi. But worse, it hadn't been scanned in from an image the same size as my print, but at one only 60% the size of mine.

So many of the in-db pics I was looking at in this section were not really larger except in dimensions -- they were all lower resolution than I could provide. Its just that the original scanner thought that bigger was better (or maybe they were trying to beat someone who'd scanned in at 300dpi earlier from the smaller 4.8x7.2 prints), I don't know. But my source was 66% larger (1/.6), and I really though they looked better at 300 dpi (which appears to be the source's dpi in both cases).

At that size, you get minimal pixelization, a slightly smaller scan size ~36-39 hundred pixels high, vs. 40-44 hunded pixels high for the in-db version, but a sharper picture (if you view the other at "80%" (83% to be more accurate - its the inverse of "200% Overscan x 60% image-size => _1.2x_ image) and my 300 dpi version side by side, you should see about the same size image. Then you'd be comparing apples to apples and could compare the actual differences in blurriness, pixelation (and color, though its size independent, mostly).

To compare the 400 dpi version I submitted (on this scan), you'd have to resize mine to 90% to see them at about the same size. Though it would be better to resize both to a 300 dpi equivalent .. i.e. view mine at 66% and theirs at 80-83%. Then you'd see the closest to what the original scan would have looked like --
Neither, IMO, are as good as the str8-up 300dpi scan, but this would require everyone realizing that bigger dimensions don't mean better quality -- and usually mean WORSE quality if the scan was done at a higher resolution than the source.

All the above assumes you are going to look at these at 100% views under PS. When you start doing that, you need to make sure all the factors are equal -- including viewing the original and look at the 'dpi' versions attached to them (they are displayed in PS, so there's no excuse not to look at them).

If you just view them side by side in two FF windows, you can just set the zooms on each until they are at the same size, or do similar in whatever other viewer one users. (MS-photo viewer, etc...).


So how do I submit a high quality scan when the definitions that different admins use do not seem to be aligned, and certainly aren't taking into account issues of color balance, source(print) dpi & size and size-size viewing of the images.

So far I've seen evidence of judging based on 1) dimension size (not good), and 2) judging based on 100% view of a scan (also not good). "1" would be correct to use if both scanners scanned off the same size image, "2" would be correct to use if both scanned at the same dpi of *source* image (which is even less likely, but possible, ie. you could have one scan @ 300dpi of a 4", and another at 400dpi of a 3", they sizes would appear the same and judging based on '2' would show the 300dpi of the 4" to be better".

But if source images and the scanned dpi are not the same, you need to adjust the viewed (compared) images to be the same visual size and compare. So 400dpi of 10", and 600dpi of 6" (4000px vs. 3600px). So lets say the wp writer wanted to use a source size of 3600 px tall, you''d reduce the larger image down to 3600, resulting in a sharper image. If *either* is more pixelated, it doesn't matter what type of post-processing (in my _limited experience_) one applies, the pixels appear as their own 'texture' in eventual scan size you use (or you blur details so much that the scan becomes of limited usefulness).

Am I making any sense to anyone, or do I appear to be talking gibberish, or am I overly concerned about quality details that most people don't care about or aren't aware of don't want to be concerned with?

Is my reality 'real' enough to be accepted, or would it take massive infusions of kool-aid to get others to notice my reality 'details'...:-)

I hope this gets seen by enough admins for their to be useful discussion -- I'm always afraid that the subject (rightfully) will indicated to them interest or disinterest).

-----

Which, BTW, was why I reposted my original request for image review under a more clear subject -- not necessarily due to impatience, but because I wondered if my subject was too unrelated to the question, AND the conversation had sufficiently strayed off the question about my the exact reasons for this particular image's deletion. I admit to having impatience, due primarily to not having much a life beyond my computer in the winter (darn cold! no gardening nor outdoor work!...Brrrr.). But I didn't post a more clear question based on my impatience, just my not being sure about not having buried my question...just...just like the turn of this question has gone on into something that I really thought deserved a posting under the genera board admin section and not as a response to a deletion appeal-explanation.

(Should I just move most of this specific stuff to another thread since I've strayed so far off (in bringing up comparison methodology) this thread's initial intent or even stated subject? :-)

Darn meandering writing style...

Real conversations don't lend themselves to nice neat tidy divisions and labels...C'est la vie.

A*s

kayurachan

Retired Moderator

kayurachan

Mado&Fen are my Gods.

Cancel

Ok, since you like to go thecnical, I'll go a bit technical too.
I downloaded again both scans, re-compared them and I have absolutely no doubts saying your scan is worse than Saikusa's.
Yes, they're scanned in a different way (400ppi VS 600ppi), so they'll result in two different version of the same picture. Yours is just slightly bigger: worst part of it it's that blur filtering you applied, plus that weird color adjustment (skin tone looks yellow...). You can see some dust here and there, too, and even your fingerprint.
You can scale them, enlarge them, do whatever you want but your scan will lack the details you can see in the other, which is not been filtered and has a nice definition.
There's not a unique rule for us mod when we need to purge a double scan. We don't always go for the bigger: we seek quality, details, nice cleaning, nice filtering and a good size (to preserve the best quality possible). I chose to delete yours since it was just inferior to the one that was previously posted.
As I already said you're most welcome to re-scan it and submit it again. It will just be a joy for our gallery to replace items with better versions.

Signature
	Image
!Death to J-Rock scans! >.<
What is Kayurachan doing? Just see it on my userpage :3
Sheq says: XD one sec, eating an egg XP

Astara

Astara

Scanner, Artistic

Cancel

Your exact details are very appreciated. The only one I'm a bit tweaked on is the 'yellow color filter'.

The original print is yellow, not red.

Many in that series have the same color shift that is off from the book I have -- I even got a 2nd copy (I cut the first one up so I could scan it better) the second one, got a few months later was the exact same shade. So if we are talking about the same thing -- I didn't apply any color filters to the scan -- the color in the 'db' (MT-db) being wrong and too dark on several of them was one of my motivations for submitting new scans.

Also, on 2nd though, not that I'm disagreeing, I'm sure finger prints are on them -- I've handled them and don't generally wear gloves. But if it was that blurry did the fingerprint and such stand out that much?

On the color cast, maybe you monitor is getting old? When monitors get older the blue part of the spectrum is the first to go. You put red&green w/o blue together and you get yellow. I recently got a color calibrator that you attach to the monitor so it can measure you monitors performance, and I consistently have my monitor set for a standard 6500k white point -- which is a press standard I'm told. I set my gamma for 2.0 -- half way between the
Apple's 1.8 Gamma and the PC's 2.2 Gamma. That way I'm not too far off either, but it's also what looks good -- the 2.2 often looks too dark (which happens to also be used in the sRGB standard), and 1.8 tends to occasionally look a little washed out. Nevertheless, I have a daylight balanced bulb for viewing and a calibrator to verify my monitor is tuned properly -- and the color I scanned in was the color as it was in the book (I didn't apply any special color correction -- my scanner just 'worked' :-)).

To my eyes, the other appeared to have been filtered because at 600 dpi the pixelization is very noticeable.

Do you prefer the scans to be of higher resolution and more pixelated, or at the resolution of the print (which was, as near as I can tell, about 300dpi)? If you apply some process to a scan to depixelate a result, what method(s) do you use?

I saw one site that recommended that recommended combining 3 copies each with different blurring types or not -- and one with an overlay. Problem I saw was that the overlay turned the scan noticeably darker and red-shifted the results.

BTW -- I am serious when I say I appreciate the detailed feedback -- it lets me know what to correct! :-)
Imprecise feedback leads to messed up corrections! :-)

Thanks for the dialog....I only want to 'do my best' as they say in so many anime...

A*a


merged: 02-04-2010 ~ 10:04pm
-------------------------------------------(...)
Next Day
===========================================(...)

Just to be clear -- on the two, side-by-side above, you think the one on the right is more clear and you like the color better there? I.e. -- you see the one on the right, above, as being more clear? Having less pixelization, and having better color?

As mentioned earlier in regards to the color -- on my monitor, the one on the right shows her skin as red even a tinge of grey in some areas. While the one on the left looks more 'yellow' (in comparison), it does more accurately represent the actual print.

What type of system do you use for viewing? Mac? PC? Size, tech & age of monitor? The idea about monitor getting 'older' came from my recent reading of the book "Color Management" -- where they said (surprising to me) -- that the useful life of a good monitor for accurate color evaluation was about 18 months. After which it becomes harder to achieve standard viewing conditions. With correction it can be extended, but eventually, the monitor will red-shift -- I think CRT's die first, but fluorescent back-lit LCD's will also suffer from the same problem as the backlight's white-reproduction fades. I'm guessing, but LCD-lit backlights should have longer lives, as the LED backlights would take a long time to fade. My LCD monitor (a Viewsonic) is about 18-20 months past its incept date and, by external measurement, is showing its age (though it looks great standalone!). It dropped from a 6500k white point at purchase to about 5900k when I first measured it with a stand-alone calibrator. I can adjust it back into range, but only by decreasing both the red and green below 50% down near 30% each, since the blue adjustment is already 'pegged' at 100%.

Anyway -- just checking that you see the image on the right, above, as more clear -- so that I "get" that my image on the left, above is actually more fuzzy. I may be seeing things with fuzz colored glasses! :-)

Thanks!
A*a

vitaamin

vitaamin

DUKE 2010 NCAA CHAMPS!

Cancel

not a scan mod but i just wanted to point out that regardless of all the technical stuff, scan modding is going to be somewhat subjective. we all have different monitors but all of us (who were consulted/saw this thread) agree that the first submission was better, mainly because of the blurring issue.

the "yellow" cast is not in reference necessarily to just the color ton, see how the lines on her leg + her skirt rag thing look really washed out? that in addition to the "dust" makes the scan less focused IMO. but i'm not really a scan mod so this is just an attempt to perhaps further explain why she is having problems with the yellow color adjustment.

Signature Image
vector-wallers tag-licious

Astara

Astara

Scanner, Artistic

Cancel

that's fine....I've lost interest on this for now...
(no need to feel like I'm beating my head against a brick wall...)
need to figure out a different approach if I want to spend the effort....

page 1 of 1 9 total items

Back to Deletion Appeals | Active Threads | Forum Index

Only members can post replies, please register.

This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies. Read more.