Darktiste Sketchbook

[IrishWhiskey]

Cool! To each their own. as long as it works for you that's great 

[darktiste]

I am working on creating a brush library that is well organize and good enough to speed up my process.Here texture i manage to do will doing so.Texture in order:wood plank, moss on a tree,waterfall.I can definitively tell there a problem with my cyclinder roundness and that water is not on point.

[Eristhe]

Hey! Looks like you're focusing on textures and composition right now, your tonal studies are looking nice. If you're looking at painting landscapes in the longer term, you're definitely going to need to integrate some perspective study into that. Likewise, for smaller objects like rocks and trees (which I see you've drawn a few of), try to deconstruct the form so that you can build them in your own scenes. Textures are great fun, but you'll need a solid base of fundamentals to really put them to good use.

[darktiste]

@Eristhe-Of course it seem i been going to fast away from construction and beaten bombarded to go back into doing more construction i just gotta kick myself in the butt in the right direction.It hard to go back when you plan other stuff but you can't skip the basic that what every drawing teach no amount of detail will fix my leak of fundamental.

On an other note i re-did the waterfall this time i actually look at a refenrence because i gotta admit water is not just blue fuck my preconception.I think this time it as improve slightly i still need to work to make the retransion from a stream of water to splash of water more fluid if you see what i mean.

My Smoke look better than what i did before it going in the right direction atleast as far as i can tell.


I also realize i should be more careful about the actual size of my canva.

[darktiste]

Here a value study i did i didn't bother posting the ref as you can see because of the error i made.Note that some of them are unfinish as i was limiting the study to 15 to 20 minute each.This is not the actual size it an error when i tried to put them all into one file it resize them into this.Let see if you can tell what i anime i am studying.I think if i take a step back one of the problem is that i still draw not enought zoom out.

I was asking myself if it was worth sharing this horrible result but i think i could get some feedback so why not.

[nigelpik]

This is Dragon Ball, right? I think, as it is, it gives you more info on how to do lighting in terms of composition (i.e., how to spread dark, medium and light values across the page), but not that much on what values to choose. Since with values, it's all about the relationships between them, I don't think using only three for such complex backgrounds makes for a very productive study. I think you should try to use at least 5 different colors for this. It would make it easier and more productive.

One thing I have to say, is that you use your whites to indicate both light and color of the object. I would choose one or the other. For example, in the first frame white shows both the light on the building and the color of the road. This makes it confusing, since it looks like the road is very shiny, which I'm sure is not the case in the original. I would personally focus on only showing light with white, core shadow with grey and cast shadow with black. If this means that the road is now indistinguishable from the grass, so be it. Because that's not the point, the point is to see how how certain objects are lit in a scene. Same can be said for the last image. White on black probably used to indicate color rather than light, but if we look at it from light perspective, it looks like there are some gloving objects in the sky.

[darktiste]

This is Dragon Ball, right? I think, as it is, it gives you more info on how to do lighting in terms of composition (i.e., how to spread dark, medium and light values across the page), but not that much on what values to choose. Since with values, it's all about the relationships between them, I don't think using only three for such complex backgrounds makes for a very productive study. I think you should try to use at least 5 different colors for this. It would make it easier and more productive.

One thing I have to say, is that you use your whites to indicate both light and color of the object. I would choose one or the other. For example, in the first frame white shows both the light on the building and the color of the road. This makes it confusing, since it looks like the road is very shiny, which I'm sure is not the case in the original. I would personally focus on only showing light with white, core shadow with grey and cast shadow with black. If this means that the road is now indistinguishable from the grass, so be it. Because that's not the point, the point is to see how how certain objects are lit in a scene. Same can be said for the last image. White on black probably used to indicate color rather than light, but if we look at it from light perspective, it looks like there are some gloving objects in the sky.

I think i forget to mention this is a 3 value study it will be more clear when i will be posting the next study anyway i am done for value at the moment i will be going into color theory now for a change because i need to mix it up sometime.I didn't felt like i learn anything about value apart from understanding that it a tricky subject to separate value in a scene and make form still read.The notan study are kinda straight forward but once you dive into 3 or 4 value pain start to begin.I think my error was mixing color temperature with value.One on my error was to try to make everything distinguishable.It new to me to think in more than 2 tone i am use to simply drawing with graphite.

+1 for guessing the anime right it was dragon ball.

[darktiste]

Limiting the time to 15 to 20 minut for a 3 value study and i still struggle to finish mapping the big value.I as you can see in the lower image i take time to separate the value but sometime i will realize mid way that i should have spent more time planning the value but since i am limiting my time i tend to spend no more than 5min planning the value.

1 slash = lightest value 2 slash =mid value 3 slash dark=value

Sometime because of leak of value to clearly separate object i make the sky dark than it is.I realized that dark sky often help separate the sky from the rest of the scene it doesn't mean that it the night.Often the sky will have white cloud so i don't want to use a mid value that whould class with mid value of the element below.

At this stage of my value study i was still experimenting on how to use the brush to block in and what tool to use.
After a step back i wonder why i just didn't use symmetry to my advantage to gain speed in the first image to create the building silhouette.

[darktiste]

I am gonna give myself a score for each of the study on 10
For how far i think there finish because as you will see most of them have been a struggle sometime the value isn't like on my value map because i had to change it to see my border.


I will substract 1 point if an big object of the scene is missing.Element that cover less than 10% of the canva doesn't count as big object.
I will substract 1 point if there only one value on a big object.



1th picture
4/10
2th picture
6/10

[JosephCow]

You don't quite have the right idea about 3 value notan. Right now you're purposefully trying to separate every element somehow using just 3 values. You're using 3 values, yes, but that's not 3 value composition. You should actually give yourself +1 point every time a big object has only one value. 

 Let me explain. 
 This is the simplest one, but it's the one you want to make the most hard. Squint. The sky is one value, the ground is another. There are a few dark values scattered in the bushes be we wont worry about that now. Those are details. They are important to the composition, but it's just not what we're doing right now. 




This one's more complex, but it's still very simple. The artist uses a lot of variations in the second value to separate objects in the scene, but it doesn't change the fact that all these shapes are the same value.




This last one is quite complicated. I think the sky must be 2 different values because the clouds separate from the roof, but the sky separates from the lit part of the house. However there is nothing in the sky that is anywhere near a dark value.  You're comparing different parts of the image to black with the little swatches I think. But you need to be comparing big shapes of value to each other. It makes no difference whether something is close to black or not, because it's all about relativity. It only matters if it's a dark in this context.





Again there are little pieces of different values scattered around, but those are details, not structural parts of the picture.


My challenge to you is to redo the one I didn't comment on and see if you can make the composition ring true.

[darktiste]

I will been posting more of them we might be able to isolate more problem there still 41 left to show.But there allready done so i won't be able to adjust until i do more of those.I get the idea of squinting i am just trying to get rid of my brain thinking it need to be a pretty picture and that i actually need to merge form.


I got a question how would you determine what the value of a big object if there more than two value? You take the most dominate value or you choose a value to separate value?

From what i understand notan are the relation of light midtone and shadow.

Thank you for the critic i appreciate.

[darktiste]

For the moment i am not gonna give myself score.But i will comment trying to remember what was the struggle i had in each piece.


1th picture
Here i think i did a ok job i have the feeling my sky should be simplfied to a simple value.Sky be white and the mountain be grey.Also the background white plain would be grey instead of white.
2th picture

Here i clearly spent to much time on specific object and i didn't use one value for each big object as suggested by JosephCow.

[JosephCow]

I got a question how would you determine what the value of a big object if there more than two value? You take the most dominate value or you choose a value to separate value?

It depends. You look for the dominant value. But if it is split into light and shadow, you certainly may make it into 2 different values. But don't make one thing multiple values just for the hell of it.

[Mylqin]

It's good to see you're studying value in such a way, getting away from details and being more efficient with my time is something most of us struggle with and starting here is a great thing. Joseph is on the money about the 3 value thing (as he should since he's such a great artist) as you're studying you have to realise that you're still juggling a couple of fundamentals in the bg shape and composition.

When you have the time you should check out Marco bucci vids specifically these 2 for now:
https://youtu.be/Nap7dwHjD9Y
https://youtu.be/-ZknWKTpc90

Also the squinting thing I learned from Chris Legaspi even though Joseph did a great job explaining it I recommend this as well, why not

https://youtu.be/zG7qa9daAbw
https://youtu.be/RxiM4uORjq8

One last thing if you can try working small... I mean you probably zoom out to make these anyway but I mean try posting them as thumbnails so we can see what you see, theres no reason to have them that large if you aren't rendering to be honest.

I can see the steady progress in your work, you're getting there keep it up!

[darktiste]

Still posting value study that doesn't take in consideration the comment to see if there anything left to fix that not allready been mentioned.

[Gliger]

Love the moon one

[darktiste]

I manage to forget 2 big shape in the first image in the lowerright corner.I really need to avoid detailling small object first.

The second image was way to complicate for the time frame i was working in.I still need to be careful about the angle of element that follow the perspective i am off on this one but i think it was just a matter of time more than anything and the study was on value not perspective so i just take a note that perspective might be slowing me atm.

I have also noted that low angle seem to be the most tricky when it come to measuring angle.I might try to rotate my canva if needed but i think i will train my eye instead.It doesn't seem to be needed since rotation can be done so fast just shortcut R and than esc.Enought with angle talk that not what i am focus on i am focus on value and working smaller that what really should matter the most.Atm.

[darktiste]

The Time frame was tight again for the complexity of the scene shown here.The first scene was a nightmare to figure out in term of value mapping.The second scene was easier to figure out in term of value but my approch.I wasted time on some aspect that gave me trouble(The spheric object.

[Amit Dutta]

A few suggestions.
If using only two values it's probably best to analyse the image and break it down in terms of lit areas vs shadow areas and blocking up the midtones into one of those much more than with a full value accurate study but without diminishing accuracy on the basic shape and placement of things. This is a common mistake I see a lot of beginners make in notan studies.  They equate 'quick study' to a licence to be really really sloppy. The time isn't important, the attention to overall comp read and general accuracy is what matters. Realise that any limited value study (<5 values)  aren't really even proper value studies. They are more about training your ability to break up any image into a simplified abstraction of the major compositional shapes and overall accurate visual relationship between the major value groupings. As such this is exactly what you need to be spending more time and attention on if you want to be getting the most out of these types of studies.   More will be learned if you know why you are studying and choose with intent where it pays off to be accurate and not shortchanging yourself for the sake of some arbitrary time.

What I see you doing here are essentially rushed looking inaccurate studies that I would argue you probably aren't learning very much from at all. 


Restricting to a set limit can be a good idea for some studies, but not if it ends up in you training sloppy, hastily done and inaccurate work for the sake of time.  In so many of these you don't seem to bother taking time or are too restricted in time to get accurate placement of the biggest shapes overall and this also tends to break the perspective a lot which is not something to ever be practicing.  Doing sloppy study work is just going to be building up lots of bad habits. And since essentially what these kind of studies are meant to be training for is the ability to simplify and achieve a quick overall read for images you create yourself without spending more time fixing things later, it's best to aim for more accuracy and nailing all those most important basic fundamentals up front. Think of these as training yourself to take the time to get the most fundamental and basic essentials of any image looking good first. If your value studies, even the basic 2 value ones, look nothing like the image being studied, you're not focusing on the right things.

Work from biggest to smallest, paint zoomed out, squint your eyes, use threshold filter on original to help you see value blocks. Use all these tricks to help you determine an overall impression rather than getting bogged up in details or being innacurate because you are trying to rush to a deadline. 

It would probably help to do analysis of the paintings before you even begin. Desaturate and colour pick the source, figure out the general value range. Try using closer values to what is in the originals for the shadow and lit sides.  it's best to have a plan and figure out in general what areas are likely to be blocked up into what value before you even start. 

Some study examples below 

Spoiler

Here's an example study I did in less than 10 mins. I consider this an incredibly sloppy study. But notice how yours got bogged down into picking shapes that really don't stand out in the image at all. All the shadow and lit areas of the forest was more easily chunked up into a single value. The mountains in the back don't even show up if you squint at the image but you made them much more prominent with a huge amount of contrast that was not necessary. Since it isn't possible to be 100% accurate with limited values, the visual relationships between the values you do use to simplify how they appear in the original, become the most important thing to actually pay much closer attention to. Contrast levels really matter.


A two value notan of one of the weakest images. More value grouping of mid tones are needed in 2 value notan. More decisions need to bea made into whether to group something into the dark or the light. Sometimes the value you choose will be opposite to the side that the value is actually closer to (ie the sky could be black with white clouds and merge into the trees, or white sky resulting in black trees which shows that contrast relationship of the horizon better, but also results in black clouds. Or the clouds could just be ignored completely and set to white since they don't really add much to this particular composition,

And lastly but potentially most importantly is your choice of study images. You seem to only be picking backgrounds from animation. Not a lot of them are particularly interesting in terms of comp and values; all are cel shaded and have very similar keys and value range and a lot rely on line to separate form. These establishing or background plates aren't designed necessarily to be looked at for very long since many also include camera moves probably..unlike standalone paintings. So I suggest you rethink the thought process behind how you choose.  I recommend instead you start branching out and finding great master paintings/illustrations and studying more from those. You can learn from anything but may as well learn from the best work you can find.

[darktiste]

In reply to Amit Dutta-

One of the problem with accuracy without mileage it that it will slow down a beginner like me and if you add a time frame to that it result into sloppyness of course there a certain amount of time where there so much time you can use that accuracy doesn't become an issue anymore.

My idea behind choosing animation is exactly that it possible to draw them under 20min compare to a master that would take much more time to analyze.

What i agree is that there no point in a rushed sloppy drawing and i think the time frame i give myself was not reasonable and i learn my leasson.I just don't want to work for hour without learning anything that why i use short study to get a quick feed back of what are the general recuring problem.I agree that one should make sure to variey the lenght of study.To see how accurate one can be.I think quick study are more about making sure you understand the concept and long study is about how far you can apply them.

Atm it just doesn't seem easy to me to be copying other work when i aspire to create my own work that why i think it result in sloppy drawing because i am more concern about applying the concept than pretty picture.I know i can be impatient sometime.I just don't want to turn myself into a unoriginal copying machine without realizing i am not actually learning anything.

You know one should be working for long period while still remaining in the realm of working with intent.It easy to go in auto pilot when you do long study.The brain reward us for learning new thing but if you do the same thing for hour it resist and want you to start over again.There that phase in a drawing where the plan as been laid and all you do is assemble the part i think we all know what is the most fun part for me it to plan i get bored when i assemble because i know how it should look.There people who love to render for hour that not my case i am more of into the idea of putting idea on paper.That why i have trouble defining how far i should push my study if after all i will be passing my work to an illustrator.

Anyway i am alway writting to much word i don't want to discredit what you saying.I am just trying to figure out if the critic apply to what i was trying to achieve.