Moderator: Community Manager
[Post Reply] [*]  Page 3 of 3  [ 30 posts ]  Go to page « 1 2 3
Author Message
Gollevainen
Post subject: Re: A problem regarding Olekit24's worksPosted: March 31st, 2023, 4:58 am
Offline
User avatar
Posts: 4685
Joined: July 27th, 2010, 5:10 am
Location: Finland
Contact: Website
This is indeed sad overall result. :( But I thank you for bringing this into all of our intention and of the work you have been done sofar.

I've set notion to the staff that we will start formulating what would be the future process of qualitycontroll. It will also had to include some level of solifying the stylerule, and it will naturally (and unfortuanetly for some) include actuall intervention against the the non-standard style choises. We are just way too big nowadays to be left on mercy of the artistist themselves understanding clearly written stuff.

_________________
Shipbucket mainsite, aka "The Archive"
New AU project "Aravala"


Top
[Profile] [Quote]
eswube
Post subject: Re: A problem regarding Olekit24's worksPosted: March 31st, 2023, 10:03 pm
Offline
Posts: 10624
Joined: June 15th, 2011, 8:31 am
@Gollevainen
Great to hear, that some clarifications to the rules and more robust quality control process are (going to be) in the making. I'm really looking forward to it, as it would make the uploading work much easier.
Of course, You can count on my willingness to provide input and help.
(but if that would mean joining Discord, I'll need a link to it - and still some time to brace myself to enter this den of monstrosity ;) )
Btw. what is the decision about these "faulty" works of Olekit? Should I delete them?

@Heuhen
Yes, the shading IS an issue with Olekit's drawing, but in regards to the "salvaged" drawings I decided to "let it be" for a time being at least.
Hopefuly, the "quality control process" mentioned by Gollevainen above will be established, and will allow to prevent repeat of such situations.


Top
[Profile] [Quote]
Gollevainen
Post subject: Re: A problem regarding Olekit24's worksPosted: April 1st, 2023, 4:56 am
Offline
User avatar
Posts: 4685
Joined: July 27th, 2010, 5:10 am
Location: Finland
Contact: Website
https://discord.gg/5PHq8Dk


Of Olekits work, I think its now best to just take down the ones not fixed.

_________________
Shipbucket mainsite, aka "The Archive"
New AU project "Aravala"


Top
[Profile] [Quote]
olekit24
Post subject: Re: A problem regarding Olekit24's worksPosted: June 4th, 2023, 8:02 am
Offline
Posts: 285
Joined: February 7th, 2016, 11:48 am
Location: Rostov-on-Don, Russia
Contact: Website
Dear friends

Finally I reached the forum, saw the message of @eswube, read the thread and, frankly, was horrified.

The comments that @eswube presented are certainly important and fair. It's sad enough that the images I posted to the forum for last 7 years turned out to be so sloppy.

The first image - MS Franko - is my first experience of creating an image according to the rules described on the forum. Naturally, it was done in terrible agony, choosing technological solutions. There were a lot of tasks that had to be solved to draw the ship in the way it was in real life. There was no knowledge, no experience, no elementary details. Hundreds of photographs and drawings, blueprints and other documents were viewed. And, of course, I did not pay due attention to the garbage that was transferred to a file with elements for rendering.

Yes, I draw in Photoshop, I actively use layers, which I then delete. This is what creates the mess that you see in the MS Franko drawing.

Since then, the technology has been honed, the elements were drawn in other files, it became possible to clean the garbarge and strange pixels. The experience came, and I learned to look where the garbage was left by applying masks.

It is strange to me to see such a number of strange pixels on the images of Mercur II, Saturn, Augusta Victoria and some others listed above. In my originals, these images are clean. Perhaps the dots appeared when transferring images to png. I use settings that assume file compression when exporting.

Although checking the image of the same MS Franko from the original file in color inversion does not show flaws, broken pixels and other things....

[ img ]

About the strange colors. Well, first of all, the basis of the image is made according to the rules: the contour is black #000000, the canvas is transparent, the main colors are white #ffffff, black, blue, red in their classic palette meaning.

For shadows and an indication of the volume, variations of white and gray are used, always the same - 2-3 shades, variations of black - 2 shades, yellow, red and others, if necessary, also 2 shades. As far as I remember, the rules do not prohibit this, but gradient. And i did not use gradients.

It was said that I use jpg elements in the drawing. Actually, not. As already mentioned, I draw in Photoshop, each element is drawn, respectively, only pixel images are present in the drawing, I do not use rasters, parts of photos or other elements that were not made by me. Over these 7 years, I have created an extensive library of elements drawn from scratch, which I use. Everything is stored in PSD, on a transparent background, which eliminates the ingress of debris or color distortion.

I disagree with @eswube, who called the situation terrible, disastrous, castastrophic, monstrous... Firstly, because an ordinary viewer will not do such difficult experiments with images that @eswube has done. The viewer will watch, use the file in his publication, print it out (all these troubles are not visible on the print). Secondly, I really don't see anything catastrophic in the situation, especially from the images over the past three years: they are quite good, there are no extra dots and the colors are the right ones.

However, here I am only a performer, I cannot and will not decide what the administration should do with my drawings. I won't decide how valuable they are to your community. Drawing them, I got a certain pleasure and realization. And the important thing for me was not to get a perfect image, but to get an image of a ship that I like. And by drawing - to study its design, structure, architecture and history. I have achieved these goals.

I suggest, since you have started monitoring my work, to make a clear list of comments and shortcomings that need to be corrected. Technical task. Then I will be able to estimate the amount of work to correct the shortcomings found and draw up a roadmap for working on errors.

And yes, more precise and detailed rules for the design of works would help to avoid such a situation in the future. It's like in life: what is not forbidden is allowed)


Top
[Profile] [Quote]
olekit24
Post subject: Re: A problem regarding Olekit24's worksPosted: June 4th, 2023, 8:19 am
Offline
Posts: 285
Joined: February 7th, 2016, 11:48 am
Location: Rostov-on-Don, Russia
Contact: Website
I will answer the attacks of @heuhen separately

Dear friend, colleague! I understand that you are a Ship and yacht Designer by profession. It's respectful and cool. I appreciate it, because you can really be considered an expert in this matter.

And I'm not offended by your criticism. But there is a problem: your comments to me in particular, are more like not friendly advice, not a response from one of the community, but the boring shouts of a boss who scolds his subordinate. I will not react to your criticism in such a tone.

You're not my boss, and I'm not your subordinate. Accordingly, I treat all your criticism - biased. So, I will not respond to your comments. But I will not, fortunately for you, do anything with the criticized image.

However, it's easy to settle, if, of course, you want to. It's simple:
- do not demand, but recommend
- do not insist, but prompt
- don't push, but help

In fact, all the images here are drawn for the common good of the community, everyone is interested in making the images as close to reality as possible.

And one more remark to you, @heuhen. Please check the facts carefully. Take the same situation with MS Hava Capella, when you claimed that only buildings were built in Turkey. Well, it's not like that, fundamentally. Turkey has completely built all four vessels, from the keel to the interior. Yes, Swedish and Norwegian specialists were engaged in the interior, but everything was done at the Turkish shipyard. And the reason for this is very simple: the vessels are leased from a Russian company, the contract involved turnkey construction. By the way, two of the four ships are still at the Turkish shipyard... sanctions, you know.

To sum up: I am calm about criticism. If the comments are constructive, like @eswube and many other forum participants, I am ready to redo, correct or supplement.

And I'm sorry @heuhen for publicly responding to you. But you started it first ;)


Top
[Profile] [Quote]
eswube
Post subject: Re: A problem regarding Olekit24's worksPosted: June 4th, 2023, 9:41 am
Offline
Posts: 10624
Joined: June 15th, 2011, 8:31 am
Great to see You back. :)
Very likely much of the issues with stray pixels and non-black-black resulted from export between various formats, and I am willing to admit, that some of it could result from You using (in good faith) parts from parts sheets that happened to be corrupted already (btw. parts sheets were reviewed in the meantime for such problems, and I am reviewing everyone's works in the archive to see for similar issues), but unfortunately these problems ARE present on Your work.

Unfortunately, as for the inverted-color sample You have provided, I'm afraid that the "testing method" You've applied would not lead to discovery of stray pixels, simply because the shade of these stray pixels is fairly similar to background white, therefore after inversion it becomes analogicaly similar to background black and stays unnoticeable... (I checked only background, without checking if the contours are right, but I believe that effects would be the same)

[ img ]

After I raised the issue with Your works, in the Discord channel started a discussion about certain peculiarities of editors that "stealthily" alter values of colors (I'm not going to quote the conclusions here, because I'm using MS Paint - and deliberately installed myself an old version - which is largely immune to this problem, and whole that discussion was a black magic to me). I'm afraid that You could fall victim to it. Still, it's up to You to provide works that are entirely clean "in the final version", I'm afraid.

As for "using jpeg" elements - I suspected that You might have had (accidentialy) export the picture to jpeg format. It happens quite often to people around, so I assumed simple explanation. That said, You also have elements on some ships (flags, logos etc.) that don't fit the SB standards (to be honest, people quite often cut corners on these) which demands such things to be pure pixelart: (these are just several samples)

[ img ]

As for my opinion that situation was "terrible, disastrous, castastrophic, monstrous..." - I'm afraid I stand by that opinion. Over 400 works were removed from the archive, because, unfortunately, Administration does not share Your view about what "ordinary viewer" would not see. Drawings simply MUST be clean (both in terms of stray pixels and of consistency of colors) - that's not up for discussion.
(btw. Your new works weren't immune to these problems. They just had much less of them)

Completely another thing (but that wasn't the reason this thread was created and your works removed, but it's an additional "problem" with Your works - though not only) is that You tend to disregard clearly stated rules regarding shading: source of light is meant to be somewhere in direction of upper-right corner, and many of your works show much of the hull being in the same shade as background white (which is bright-est shade of the scale), while the front/up facing parts should be even "whiter" (that applies to ships in other colors analogically).
(generally, black should be used on contours, never as "paint", and white on background, and also never as "paint", though esp. older works unfortunately use "background white" as "brightest shade of white paint", so it's kinda tolerated, but not the intended way)
Also, way to often the portholes on Your works are NOT outlined in black, while they should be.
And last but not least, there's an issue of hull shading. It's common error, so it's not disqualifying works from being uploaded (because the comprehensive hull shading guideline was never formulated - but it will be), but it's not shaded as it's intended to be in the SB style (simply speaking, You shaded it based on diagrams showing the hull "cut" length-wise, whereas it ought to be based on diagrams showing the hull "cut" width-wise).

I really hope there would be some way to fix Your works, because they are potentially (in a "correct" condition, and perhaps with some factual errors that might or might not be there sorted out) very valuable input to the Shipbucket.


Top
[Profile] [Quote]
olekit24
Post subject: Re: A problem regarding Olekit24's worksPosted: June 4th, 2023, 10:50 am
Offline
Posts: 285
Joined: February 7th, 2016, 11:48 am
Location: Rostov-on-Don, Russia
Contact: Website
Well, what to say... the flags and emblems given as an example... Yes, at first I used flags and emblems from photos, however, after getting scolded several times by colleagues on the forum, I stopped, began to draw them from scratch. I also used a template posted on the forum with flags.

The HAL and Dobroflot emblems are drawn by me. I see that I haven't cleaned the pixels. This can be fixed.

I didn't really understand about the shadows. Yes, I draw a side view, respectively, I indicate the shadows on the body in accordance with the side projection of the body. Without shadows, it would just be a painted drawing, wouldn't it? I don't understand the idea of "drawing shadows relative to the width of the body", I draw the body in length, it doesn't show its volume anyway.

Regarding portholes that are not circled in black: yes, there was such a period. But I also went back to the rules - I draw in black. You did not support my idea that the contour of the aluminum frame cannot be black at the time. Also, I don't draw the rails in black, since in one case it's just a steel rope, and in the other it's a thin iron structure with wooden railings. There seemed to be no complaints on your part.

Damn weird pixels. Well, I can't guess where they come from, if three-dimensional colored backgrounds, such as the underwater part, for example, are made by filling, and on your inversions it looks as if it was stuck with a pencil.... I'll check my sources.

400 illustrations that I have drawn for 7 years will not be fixed quickly. But I appreciate your appreciation of my contribution to Shipbucket. I'll try to think of something.


Top
[Profile] [Quote]
olekit24
Post subject: Re: A problem regarding Olekit24's worksPosted: June 4th, 2023, 10:58 am
Offline
Posts: 285
Joined: February 7th, 2016, 11:48 am
Location: Rostov-on-Don, Russia
Contact: Website
By the way, eswube

Can yo show me the example of right shaded and with right colours image. I'll explore it.


Top
[Profile] [Quote]
eswube
Post subject: Re: A problem regarding Olekit24's worksPosted: June 4th, 2023, 11:48 am
Offline
Posts: 10624
Joined: June 15th, 2011, 8:31 am
Although the Style Guide (http://shipbucket.com/styleguide) needs some clarification and update, it's still valid source and it provides at least some answers to Your questions.
As for railings: http://shipbucket.com/styleguide#33
As for general pattern of shading: http://shipbucket.com/styleguide#5

Below is a sample of more-or-less correctly shaded ship (certain things are subject to personal preferences and interpretations).
(Yellow means highlight - areas on which light shines, orange is "base", red is shade, maroon is dark shade and purple is contour insufficient to be black)
Also there are some simple diagrams to show the general idea of shading in SB.
And lowest is the sample of hull shape diagram (of a completely unrelated ship) - in blue I marked the lines You've been usually using (from analogical diagrams) to draw shading of hulls, in red I marked what You should use, and in green approximate area that would be shaded.

[ img ]


Top
[Profile] [Quote]
olekit24
Post subject: Re: A problem regarding Olekit24's worksPosted: June 4th, 2023, 2:39 pm
Offline
Posts: 285
Joined: February 7th, 2016, 11:48 am
Location: Rostov-on-Don, Russia
Contact: Website
The buttock lines in the theoretical drawing of the ship were just invented in order to show the curves of the ship's hull in a flat drawing. Accordingly, and this is an indisputable fact, the part of the body below the buttock line is narrower then the part that is above this line. So, with the frontal lighting that you are talking about, the shadows are distributed along these lines. And the closer to the waterline, the narrower the hull and the darker the shadow.

However, in this case, with frontal illumination of the object, the stern lines of bends will not be so obvious. I agree here.

It turns out that the shadows will look like this: ||\ ///, where \ and / is shapes of buttock lines.

That is, in the stern, the shadow is aligned with the width of the hull, and in the bow - based on the contours of the hull and different widths at different heights.

Otherwise, I agree with most of you explanations.

Thank you so much for the illustrative example.


Top
[Profile] [Quote]
Display: Sort by: Direction:
[Post Reply]  Page 3 of 3  [ 30 posts ]  Return to “General Discussion” | Go to page « 1 2 3

Jump to: 

Who is online

Users browsing this forum: No registered users and 8 guests


The team | Delete all board cookies | All times are UTC


cron
Powered by phpBB® Forum Software © phpBB Limited
[ GZIP: Off ]