Anonymized Measurement Database


#1

Given the insane number of clothing size standards and how long ago most measurement surveys are (The most recent I’ve found (SizeUK) is from 2004), it just struck me that it might be VERY handy down the road to request permission from users to upload anonymized personal measurements for a public database so that measurement relationships can be established for the desired end of creating a multisize formula for default measurements and grading. (For example, people with a height of A, a waist height of B, will usually have a crotch height of C and a knee height of D.) Filling in more values would create more and more precise suggested values for measurements that haven’t been filled in yet. Ideally, simply entering someone’s standard sizes and a few essential measurements would create a measurement file that would fit 99% of users.

I’d say that this would require some pretty sophisticated statistical coding, but I could see it being a huge boon to a lot of people.


#2

The problem I see right now is that the database will not be representative, as it’s mostly people who don’t fit existing sizing systems who try custom clothes


#3

And I don’t really believe in 99%, however good the database will be. People’s bodies are just too different. That’s why the more presize patternmaking system is the more measurements is required.


#4

The database wouldn’t be based on set sizes, but actual measurements, and the relationship between them. Allowing users to input their sizes would just be a shorthand to fill the measurement table with values that are somewhat close. As each measurement would be added, the other values would change based on their relationship with the entered value.

For example, let’s say that someone said that they were an ‘XL’, which, according to Aldrich, means that their expected bust, waist, and hip size would be 106, 90, and 114, respectively. If that person then entered 102 for the bust size, the other two probably wouldn’t change much, but if she entered 98 for her waist, the hip measurement might automatically become 117. Note that this change wouldn’t be based on her ‘size’, but the calculated relationship between the waist and the hip, which in turn is based primarily on the relationship between the waist and hips (and, secondarily, of the other measurement relationships) of THOUSANDS of other users, plotted on a graph and turned into a formula using curve fitting.


#5

Good idea,

I suspect that length measurements especially would show a useful correlation. For instance forearm length as a ratio of arm length. I’ll add getting this permission to the measurements entry area of the pattern share which I am revamping currently.

Jase


#6

In an effort to understand measurements and proportions of the human body to TRY and find the ideal way of creating a standard set of measurements, I’ve been studying through various methods and systems. I started with The Practical Designer (1918) by Prof J Rosenfeld, but this system is difficult to translate into measurements for a multisize Tape file.

Since then, I have looked at da Vinci’s Vitruvian Man but once again there are some discrepancies but it did point me towards looking at proportions in drawing figures which uses the measurements of the head to establish the proportions of the rest of the body - Human Figure Drawing Proportions - perhaps this will show you.

However, my studies in this regard continue… :slight_smile:


#7

Haha, you know, it seems like standart sizes template in Valentina was created specially for (post)soviet state standarts, nothing else

Btw, those things have like about 200 different sizes and still don’t fit many people, that’s part of why I’m sceptical of any statistical method


#8

I totally agree that any statistical method won’t fit many people but it would be nice to have a statistical basis on which to work that one could edit the formulas to specific measurements without having to go & create a new measurement chart for each person from scratch :slight_smile:


#9

One of my favorite things about Valentina is the ability to create individual measurement files. I have seen many standard size sets and found that NONE of them is reliable. When I lived with my 3 sisters and mother in the 1970s (5 women in one house) we ordered a garment from a TV commercial that was advertised as “one size fits all”. ONE SIZE FIT NONE.

The first project I sent to a 3d printing hub was to print these models

I take these with me to illustrate a point when someone wants to know why it is complicated to dress a large group of actors alike and to make them all look good.


#10

I still can’t understand what’s the use of the proposed system. If that’s not like standard size table, but a suggestion, then you will need to check the suggested value if you want a good fit. And if you don’t, why would you bother with custom pattern at all


#11

If you read the introduction to The End of Average (intro is the ‘Look Inside’ section of the book, free to read), you’ll find out that this is a known problem. Very few people (less than 20%) actually fit the average size, no matter which population you pull the data from, no matter how new or old the data. The concept of average sizes itself is flawed. Read the intro chapter to get the extended description of why this is true.