Artists’ Workshops: Stryjeńska // Ewa Oberska Capstone

Artist’s Workshops: Stryjeńska

Fine Arts is an industry that exhibits a significant gender disparity in the representation of artists. Currently, it is estimated that more than half of American and European visual artists are female, yet the presence and recognition of their works cannot be compared to that of their male counterparts. The imbalance not only misrepresents the presence and contribution females to the art industry, but also provides a false impression of inferiority of their works of art.

The project strives to combat this disparity through providing the audience with an interactive space that educates on the works and lives of lesser-known female artists. It focuses on artists whose substantial contributions to the art scene do not receive adequate recognition in the industry and in media.

As Artists’ Workshops: Stryjeńska is the first part of this project, it focuses on Zofia Stryjeńska, an artist of the Interwar period in Poland. Her story of becoming one of the most influential European artists of the time is that of struggle and constant fight against social limitations imposed on women artists. The obstacles Stryjeńska overcame are representative of hardships that other female artists faced which made her a perfect fit for the focus of the experience.

Stryjeńska’s workshop is a navigable VR space that the audience is free to enter and explore. The environment is equipped with a number of 3D objects which upon interaction, reveal the story related to them. The items placed around the environment include paintings, drawings, but also letters from her son, sketches, notes, personal calendar, and more, which serves to introduce the artist’s professional work, but also her personality, the deeper story behind her success.

In order to provide the audience with historically accurate information, the items, as well as the storytelling triggered by them, is based on the artist’s personal property documented by Polish National Archives, as well as her own diaries written in the 1920s and 1930s.

As Artists’ Workshops is a project still in progress, more workshops of other unrecognized female artists will be developed.

PPT presentation:Ewas-Capstone-Final-PPT

Video documentation:

mis(s)representation – final project // Ewa Oberska

Mis(s)representation is a visual project that aims at bringing attention to the different ways in which males and females are represented across visuals in advertising campaigns and throughout popular media.

The project consists of two parts – the main one which comprises of a series of combined images where gender roles have been switched, and the second part that shows the original images as a point of reference and a way to emphasize how differently we perceive female and male roles.

The Process

Research: The process towards creating the series started by conducting a vast research online on the problem of sexism in advertising and popular media. The country that truly stood out in terms of the ‘casual’ sexism featured in media turned out to be the USA. Therefore, all original images come from American websites and focus on the image of women in the American culture.
Two specific companies that I found to be ‘outstanding’ in how they systematically represent women as submissive and as sexual objects are American Apparel and Tom Ford, both being fashion companies. I made sure to include images coming from these companies in my project.

Pre-manufacturing & gathering assets: Gathering sexist visuals of women in American media was only a part of my research, in fact, the easiest one. In order to Photoshop these images to switch the gender roles that they represent I also needed to collect photographs of men in similar poses as the depicted women. Needless to say, it was an extremely time-absorbing process, as there are not nearly as many images of naked men in explicit bodily poses available online as those of women. The solution I finally found was that in order to ‘build’ one male body, I would download multiple images of males and combine their body parts to make the ‘final man’. In some cases, I had to use as many as 5 different images of different males to make one, comparable to his female counterpart.

Personal image use: Due to the nature of my project, before I started creating it, I decided not to use randomly found images of males found online. Original images are a great example of sexism and diminishing the role of women, found by many offensive. Therefore, since I had no permission to use the models’ images, especially to depict them in a very poor light, I wanted to avoid offending any of the males whose images I used. For that reason, every image of a face in my project consists of at least 2-3 photographs of different faces, photoshopped and combined together, so that the identities of the models are not recognizable.

Image manipulation techniques: To achieve my final effect, I worked in Adobe Photoshop. The goal was to make the ‘remixed’ images as similar to the originals as possible. That goal was sometimes very hard to achieve given the difficulties in finding proper images to photoshop, due to lighting used in original images which was hard to re-create in Photoshop, etc. Apart from merely combining multiple images together, all altered pieces also required some touch-ups, which I did using a drawing tablet. In a few cases, instead of importing an image of, for instance, a male arm, I made the male arm through blurring, smudging and stretching tools (mostly), and drawing the shadows on top of it, etc.

Presentation & user testing: Due to the nature of the project and the fact that it is strictly based on images already existing in media, the way in which the project is supposed to be presented is a crucial part of it and has a significant influence on how it is perceived by the audience.
Through user testing, I found that first showing the original images and then the remade ones the effect was quite unimpressive. The reaction I would get was ‘well yes, it’s a problem’. On the other hand, I started showing only the remixed images which invoked much stronger feelings. Without the point of reference,  without providing the originals, the audience was much more pushed towards understanding the purpose of the project on their own, analyzing it critically on their own. In other words, the project was much more surprising and stimulating to the public. Many of my viewers, however, mentioned that they wished that after viewing the remixed images, they could see the originals and compare them. Seeing their reactions I was satisfied to notice that they were actually actively analyzing the images, and the surprise effect was much larger than if they were shown the originals first. Therefore, the project primarily consists of the remade images, with an option to view the originals.

The Project

Here are some of the images featured in my project:

Original images:

User Testing Feedback #2

This week I conducted multiple user testing sessions with people from outside of my academic environment: Amanda, Jessica, and Cris. Considering I am already quite far in the manufacturing process, the feedback I received was much more specific than the first time around.

My most important question to the testers was whether the purpose and story introduced by the 3D experience is transparent and understandable, or whether it needs to be slightly altered to convey its purpose more clearly.  The question I got asked by all of my testers, in the very beginning after they entered the experience, was ‘where am I?’ and ‘what am I supposed to do here?’, which made me realize that I need to provide the user with some simple and minimal instructions, as well as an introduction to the environment and the character whose life it’s supposed to portray. Especially considering the fact that the subject of my project, Zofia Stryjeńska, is not a very well-known artist. Apart from that, they also had some trouble understanding how to move around in the environment, which I realize is the fault of the locomotion that I am still working on.

After finishing the environment, all of my testers mentioned to me that aesthetically, the experience was very satisfying, but not too educational, as there were only 2 objects they could interact with. I definitely intend to create more interactions in the environment attached to 3D models that I have already created.

Apart from that, something that my testers noticed, that I had not, was that when you actually enter the experience using a headset, sometimes a number of objects start flickering. I am not sure why that happens, possibly because of the lighting, so that is something I still need to work on. My testers also mentioned that it would be good if there was more to explore, if there was a number of rooms they could go to to learn more about to artists. However, realistically, I do not think I will be able to accomplish that by the time the project needs to be presentable, so for now I will stick to what I have, and after the semester ends, I will keep working on the project to make it into something, hopefully, truly influential and publishable.

Visual Remix Artist: Penelope Umbrico // Ewa

I came across Penelope Umbrico’s work as I was reading a New York Times article on the idea of the visual remix. I viewed the works of all mentioned by NYT artists, but I found her work to be the most appealing.

The first of Umbrico’s pieces I have seen was 541,795 Suns from Sunsets from Flickr, which was also made the cover image for the NYT article, and I believe it was made the cover image not without the reason. It is a collage of sunsets which I found to be impressive, especially after seeing a photograph of the piece exhibited in a gallery. The artist looked for the most frequently photographed subjects online and found them to be the sun. That drove her towards creating the collage. I believe that an extremely effective technique that Umbrico used is how she chose the images. Every image has a high contrast, saturation, and very defined sun. She also cropped the pictures in a way that eliminates all other elements from the photographs and makes the sun(s) the sole focus of the images.

The second piece I studied is, in fact, the artist’s response to the first one. Penelope Umbrico created Copyrighted Suns/ Screengrabs in order to “question the claim of ownership of an image of something that is essentially un-possess-able”, as she describes it on her website. As a statement, she removed all actual suns from images and replaced them with watermarks that symbolize copyrights.

What both of the pieces have in common is that they are fully obtained from an outside source (Flickr), and combined together to create this beautiful outcome. They are created using very similar techniques – for instance, the ‘sun’ subject in the center, high saturation. The technique is, in fact, so similar that looking from afar, one could not distinguish between which piece features sun, and which one features watermarks.

While I find the first piece to be extremely appealing, probably one of my favorite pieces of visual remix art so far, I do not find the second piece not any less engaging. Even though visually, I favor 541,795 Suns from Sunsets from Flickr, I am impressed by the idea behind Copyrighted Suns/ Screengrabs. I find it to be an extremely smart and ingenious way of addressing the case of ‘owning’ images, which the artist admits she does not agree with. Moreover, through this work, Umbrico also connects to her first piece and makes a statement about the source images she used to create it.


Twitter Bots – Think Piece // Ewa

In my Twitter research, I found three text-based Twitter bots: ‘@TwoHeadlines’, ‘@DeepDrumpf’, and ‘@kookyscrit’. The mentioned bots work in very different ways and obtain their original; text from various sources.

TwoHeadlines is a truly amusing (and funny) bot that bases off of Google News, from which it takes popular headlines and remixes them creating text mashups of a kind. The outcome of the process is yet another headline. As the bot does not necessarily choose headlines that are relevant to each other in any way, the outcomes are often very humorous.

Another bot that I found to be very interesting is DeepDrumpf. After doing some research, it turned out that the idea of DeepDrumpf actually originated from John Oliver’s show ‘Last Week Tonight’. The bot obtains its source material from Trump’s speech transcripts, his own tweets What I find to be a wonderful aspect of this twitter bot is that it serves to fundraise for girls in the STEM field. Bradley Hayes, the creator of DeepDrumpf, in a press release, admitted that Donald Trump was a great speaker to base the bot on, considering his speech patterns, which I completely agree with. The bot-created tweets resemble those of Donald Trump himself in a number of ways – in terms of his use of language, grammar, and combining a number or unrelated to each other sentences. I believe that a large portion of the population, if told that DeepDrumpf  tweets are actually Donald Trump’s tweets, would be likely to believe it.

And for reference:

The last Twitter bot I researched was one similar to ‘GrammarPolice’ – @kookyscrit. It focuses on the use of 2 words: ‘weird’ and ‘definitely’. From what I have seen, the bot posts approximately every 10 minutes, targeting other Twitter users who made a mistake in one of those two words. What I find amusing is that the bot has actually more than one way, pattern of responding to misspelling Twitter users. For instance:

  • “It’s an ‘i’, not an ‘a’: definitely”
  • I before E except after C: weird
  • Did you mean to say “Weird”?
  • That’s a kooky spelling! I think you mean “Weird”

Something that I noticed and find funny is that the bot itself makes a mistake (‘Its’ instead of ‘It’s)’

I think the reason Twitter is an extremely effective platform, especially for text-based bots, is that it’s very accessible and open. Bots have access to people’s tweets while, in the case of Facebook, most users’ privacy settings would hinder the process of obtaining data (posts). Moreover, Twitter is very text-based. In the case of Facebook, I barely ever see posts that contain text only. Having done research on video and images attached to posts on social media, I know that the audience of text-only content is nowhere near that of posts featuring some kind of visuals. That is the case especially present on Facebook. Twitter is perceived slightly differently and Tweets are in general shorter as well, which I believe makes it a platform where text-only content is much more successful. Hence, I think it makes a great environment for text-based bots.

Final Individual VR Project // Ewa Oberska

  Idea: My dream painting workshop in 3D
Technology: Maya 3D (3D modeling), Autodesk Mudbox (model texturing), Adobe Photoshop (texture creation/editing), Unreal Engine (3D environment, interaction).

Working on the individual project alongside the group one, it was both helpful and slightly challenging. Helpful, because I did not need to learn one thing twice. Challenging, as I focused my efforts mostly on completing and polishing the group project. I decided to make a ‘dream painting workshop’ as I thought it would give me the chance to basically create a ‘pretty’ environment. I also imagined it to be inside of a beautiful, Victorian house in the middle of nowhere. At this stage of my project, I focused on creating the workshop itself, not the house, but I am planning to keep working on it and further build other locations in the environment. The reason why I wanted to place the workshop as I said, ‘in the middle of nowhere’ is that I have always been attracted to the aesthetics of beautiful natural environments in 3D and thought this would give me a chance to work on one. Besides, what artist wouldn’t like a peaceful and beautiful environment to create their works in?

In terms of assets, my initial plan was to create all of them. Towards the end of our 7-week semester, I already had created multiple 3D models for both the group and individual projects, but I had to let go of the ambition and download some ready 3D models, at least until I would be able to make my own. Here are some of some of my models created for this project:

I think being able to create my own assets and texture them is a great advantage, as I am able to make my environment exactly the way I want it. No need to compromise because ‘that was the only free 3D model I could find online’.

In terms of the sound, I downloaded some natural forest from and attached it to the environment outside.

As for textures, I mostly used google images, made sure to choose files that are as large as possible to ensure the quality of my textures and a real feel to my environment, which wouldn’t be possible if the textures on my models were pixelated. Some textures, I had to edit in Photoshop, for instance, editing the hue to achieve a certain color that I wanted or to simply copy the image multiple times to make the texture seem more detailed.
I had exceptionally good time when working on glass texture. Once I figured out how to add different hues to the glass, change the opacity, amount of light and image that can get through the glass, it made my environment look quite authentic.

As for locomotion, it took me some time to figure it out, but considering I had already gone through the locomotion part with my teammates it was not nearly as time-consuming as the first time. I tried out teleportation, but it seemed to me a little unnatural. Therefore, I decided to use sliding movement in my project. After slowing it down a bit, I decided it was a right decision.

In terms of lighting, I spent some time trying to make it look like natural morning light enters the room, but at some time I accidentally did something and the lights weren’t working anymore. I tried to get in back in multiple ways, but probably made it even worse. In the end, I did the lighting again, however, with a worse outcome simply because of time constraints. I’m planning to keep working on my project though so I’ll give natural lighting another try.

I consider my project to be a work in progress. I want to polish it visually, slowly create all missing 3D models and delete the ones I downloaded, and lastly, add quite a bit of interaction. I would like to make this project something I can show as an example of my 3D design skills, but to achieve that, it still requires a vast amount of work. As for the learning outcome, even though the group project was my main area of focus, I know I learned a lot from individual work as well.

Capstone Project Update // Ewa Oberska (Rune)

Over the past few weeks, I have been constantly working on my Capstone project – Digital Museum of Female Art. Despite the evolution of my idea, the name of the project still remains the same, however, I’m considering adjusting it to what my project is becoming.

I have been working on my project through completing small tasks that led me to create a ‘draft’ of my final project.

  1. The story, research of the artist: I decided to base my work on the story and art of Zofia Stryjeńska, Polish painter who disguised as a man to be allowed to attend the prestigious Academy of Fine Arts in Munich, and remained loyal to her art and expression, even during the hardships of war. The research involved reading her biography written by Angelika Kuźniak and doing research on how typical art workshops looked like in the time when she created her pieces.
  2. Assets: I started planning out my project and decided that there is no point creating environment if I have no plan for it, no assets, anything. Therefore, I started off with creating 3D models for my environment that would match the style of the time period. Even though it was quite a time-consuming process to create the models, it was worth it. My digital sculptures are consistent with each other, and are exactly what I wanted. An easier way would have been to download the 3D models, but I did not want to do that for 2 reasons: I want my capstone project to be a representation of my 3D design skills. The second reason was that no website offered me the exact models I imagined for my project. Then why compromise if I can make exactly what I need?
  3. Textures: I have downloaded textured from various websites, all of them are large files in order to make sure the texture quality is good and that there isn’t any pixelation, for example, on the walls. I also had to adjust a few 3D models to accommodate my textures. For instance, in models where polygons were not evenly distributed, my textures would in some places look very stretched and in others very dense. Some of it I fixed in Maya through editing the models themselves, but for others, I simply edited the texture settings in Unreal Engine. I also had to edit most of them in Photoshop to better match the needs of my project.
  4. Environment: I have built a simple environment. An art workshop, with a few windows, balcony. The user will not be able to exit the room however, I am still going to build an outside environment (a natural one). I think it will be of a great aesthetic advantage to the environment.
  5. Locomotion: I managed to create some motion in my project. It still needs a lot of polishing. I believe it’s best to work one step at a time, so my main purpose at the moment is to finish creating the environment and then I will be able to move on to interactions and locomotion.

Overal: The project is still quite rough. There are a few things I plan to work on:

  • The character – the painter herself, if I decide to involve her in the scene. That would need me to model and animate her which in itself, would make up probably more than half of my time spent on the manufacturing of the entire project. To model a realistic humanoid, texture her, work on her clothes, fabric and then rig it to be able to animate is a huge undertaking. It has to be extremely well done, otherwise it will simply look ridiculous and be a disadvantage to the entire project. Therefore, I am still thinking whether I should try to do it, or just focus on creating an impressive experience.
  • The lighting – currently it leaves much to be desired. I definitely need to spend at least a few extra hours making sure it looks natural and adjusting the placement of my models so that there aren’t any random shadows where they should not be.
  • Some other models and textures – not everything looks perfect just yet, there are a few models I still need to create and textures I have to change or edit. I also need to generate and add bump maps to my textures to add depth to the environment.
  • Locomotion – certainly needs at least a few hours of work. Currently, the player moves a little too fast in my opinion.
  • Interactions – I still have not touched interactions just yet. I have been researching how to make them in Unreal Engine, and I have a few ideas. However, all interactions will be attached to specific models therefore, I need to finalize the contents of my environment before starting work on the interactive part of the project.
  • Audio – will be a crucial part of my project. I have created an audio map and planned out which directions the sounds will originate from. My work on audio will only be possible after I have finished implementing interactions in the environment. Also, I will need to record a little bit of a monologue, but that should not be a problem. Initially, I intended to find a native English speaker to record their voice, but I later came to the conclusion that I will be better off doing it myself, sticking to my own accent. After all, the painter was Polish so it might actually add to the authenticity.

Video Manipulation as a Political Tool – Remix Midterm (Roopa)

For this project, I decided to experiment with video manipulation and remix on my own. As my midterm paper discussed various methods used by Hillary for America in their video ads for the 2016 Clinton’s presidential candidacy, I chose to further research the field.

In the case of my project, I opted to utilize videos of different Donald Trump’s statements, especially those made during his presidential campaign. The purpose of this project is not to convince the audience that Donald Trump is either wonderful or despicable candidate, however, it serves to show how context, material manipulation (such as the choice of color gamma, edits made to videos), music are capable of completely changing the conclusions that the viewers may draw from given statements.

In order to convey my thought and purpose, I created two contrasting videos: one pro-Trump, and one anti-Trump. However, the resources used in both videos are exactly the same. I made that choice to make my point even stronger and more convincing. On the other hand, I also chose the most controversial statements of the Republican nominee, such as “I will bomb the sh** out of them”, “we have people that are stupid”, “American dream is dead”, etc. That made the project significantly more challenging. How do I convince my audience that “I will bomb the sh** out of them” is a positive statement, one of the reasons they should see D. Trump in a good light? I used a number of different methods, some of them being barely noticeable, especially if one watches the videos only once, however, I do believe they impact the viewers’ subconscious perception to some extent.

  1. The Process:
    Prior to creating the videos, they required me to do a large amount of research. I spent hours listening to Donald Trump’s speeches and noting down quotes, statements of his that I thought were generally perceived to be extremely negative/violent/hostile, that I wanted to completely turn around and make them seem to be a good thing.
    For the purpose of my research paper, I researched the data on the most successful videos on social media and learned that the most watched in full and popular videos are 1-2 minutes long which is why I decided to keep both of my videos in that time range. Therefore, out of almost 60 statements that I had previously written down, I had to choose the best 6-8 quotes.
  2. Having chosen the 7 most suitable video clips, I needed to come up with the arguments that they would advocate for: both positive and negative. In order to ease the process, I made a table where I compiled all statements and the narrative that I wanted them to convey in both videos. Over time, the table went through a number of changes but its final version is more less as follows:
  3. After my idea and action-plan were set, I began the process of video editing.

Methods used to manipulate the perception of chosen statements and recontextualize them:

  • Subtitles: they played a large role in both videos. They are certainly the most noticeable and direct method. They leave no room for speculation, instead, they provide the audience with ready conclusions on the statements. For instance, showing one of Trump’s most known opinions on relations with Mexico, the video comments “hostile“, which automatically tells the viewers what to think, before they even manage to think critically themselves. It certainly does not mean that whoever watches the videos will automatically agree with the subtitles however, it does provide the conclusions basically ‘on a golden plate’, which I consider to be a very effective method as human minds can be lazy and go for the easier option which, in this case, is merely absorbing the provided information.
  • music: initially, I downloaded a video from AJ+ facebook account and pulled the music out of it. The video was about a mother from Missouri who was a candidate for a senator and has done a lot of good for the sake of the issue of breastfeeding in public. Of course, the accompanying background music was very cheerful so I decided to use it for the ‘pro-Trump’ video. After editing the music and cutting out baby sounds out it, I put it together with some footage, which is when I realized that the music was too cheerful. 2 hours later, I found what I consider to be a perfect audio background for the ad.
    AJ+ is known for very short, yet impactful videos, often of political nature, that feature very little or no narration, which in my case was perfect. Therefore, I did end up utilizing another background music downloaded from the AJ+ facebook page for the anti-Trump ad. The original video created by AJ+ that used this audio was about the abuse of animals in Indonesian zoo’s, but surprisingly it perfectly matched my editing needs.
  • Color gamma: I changed the main colors of both videos: anti-Trump having a black background, and pro-Trump with a white one. Upon comparing those two, I thought the difference is quite noticeable.
  • Fonts! Both of my Trump ads feature a completely different choice of fonts. The video condemning the Republican nominee uses an especially heavy and thick font, being Arial Black, to emphasize the weight and seriousness of his statements. However, in the case of the pro-Trump one, the font is significantly lighter, even animated to resemble being handwritten, which is more commonly used
  • Donald Trump’s images: I also edited Donald Trump himself! For instance, in the ad that shows him in a better light, I showed him with family, smiling, giving off a more positive and peaceful feeling. Moreover, I also edited his visuals a little to make him appear slightly taller and skinnier.

4. Videos: The two videos below are the result of my work. As there are countless videos online featuring the statements I used to criticize Donald Trump, I made the anti-Trump video merely to make it a comparison point for the second one. Therefore, I believe the pro-Trump ad is much more technically polished as it makes the main component of my project. // (I tried embedding his video, but for some reason wordpress does not allow me to do so)

What would I have done differently? I think if I were to approach this project the second time, I would choose different statements made by Donald Trump. I believe the ones I chose this time work well however, it would be easier and probably a little more effective if I chose statements that are easier to perceive in both a positive and negative way. This choice certainly made it challenging to show Trump in a good light. Moreover, I would also make the videos a little more animated to polish the visual effect. I believe the pro-Trump ad works, but the anti-Trump one could benefit from some smooth animation effects.

Capstone User Testing #1 (Rune) // Ewa Oberska

Questions asked:

  • Can you identify what kind of an environment/room this is? What purpose does it serve?
      1. “Some kind of a study with a bunch of easels.”
      2. “A painter’s workshop! I mean I see all the easels and paint all over the place.”
      3. “Definitely an art room, probably belonging to a female artist.”
  • What did you think when walking through the hallway?
      1. “Confused.”
      2. “Curious.”
      3. “Kind of gloomy, didn’t give off the ‘museum’ vibe.”
  • Your thoughts on this project: should it be in the form of a game or VR?
      1. “VR.”
      2. “VR.”
      3. “A game, VR requires a headset which not everybody has or likes to use.”
  • Did you feel intrigued by the environment, feeling like you would like to learn more about the person who it belongs to?
      1. “Kind of, but then I was pretty confused.”
      2. “I realize the project is far from being completed, but despite that I’d like to see it’s final version and learn why you chose to describe a person through their own space.”
      3. “Absolutely, I’m very much into the idea of female art so I’d love to learn the story behind the artist you chose for your project.”
  • Do you think this environment would benefit from having a protagonist, a modeled and animated character in it?
      1. “Yes.”
      2. “No difference.”
      3. “Not necessarily. It could be helpful if it’s really well made, if you can make it possible to interact with that character OR if she plays some role in the environment. If you just put her up there like a statue, it won’t be very helpful.”
  • Anything you think I should particularly focus on?
    1. “The story/purpose. At this point, it’s not very clear.”
    2. “I think you could put some work into texturing and visuals. And definitely interactivity.”
    3. “The idea sounds great, but I think you could benefit from focusing on the interactions and conveying the purpose of the project a bit more.”


  1. Definitely need to work on the visuals to avoid ‘confusion’, in particular, the initial ‘hallway’.
  2. Re-think whether I should stick to the ‘game’ form, or maybe allow my users to experience my project in VR.
  3. Especially focus on the storytelling and interactions!

Heavenly Castle Fish – 3D models // Ewa, Zeerak, De’yon

We have a few models created specifically for the purpose of our VR project:

  1. A desk:
  2. A wardrobe, decorated in the same way as the desk in order to keep the style of the environment consistent:
  3. Balcony entrance/arches – also decorated with the same patterns as the two previous models, however, in this case, it’s slightly less visible:
  4. And finally, balcony railings!
  5. We also have a few other models, like stone steps, door arches, etc, however, they’re not as presentable as the ones above. We will soon also have a self-made chandelier!