Basics

[theropod]

Something slightly different this time. Not strictly the kind of tutorial I expected would be needed here, but this is the basics thread, so if a tutorial is needed, this is where it goes.

Making size comparisons
I have noticed size comparisons I have seen here are often inaccurately scaled, poor resolution, or unnecessarily using bad images that appear to be easier to scale, or some combination of the these. So I thought I would give a short tutorial how to properly do them, using only free software.

1) First of all, use accurate measurements. They should be:
A) Sufficiently large relative to what you are scaling (don’t use the length of a single tooth of vertebra if you can use shoulder height or total length)
B) reliable (from a reliable source, properly measured and not guessed)
C) verifiable (you need to be able to identify said source)

I think no further explanation is necessary, as everyone here should be aware what does and does not constitute reliable data. The most important thing is to properly state what dimensions you scaled your animals to, not just some vague statement that people won’t be able to check. I.e., if you scaled something to "average size", state what constitutes said average size, and describe or link how it was measured or estimated.

2) It should go without saying that before proceeding, you need images that:
A) you are allowed to use AND
B) are accurate AND
C) you can accurately measure the measurement that you are scaling them on in.

For these reasons, skeletal reconstructions ("digital X-Rays") are often preferred, as they facilitate the measurement and accuracy-parts. However, a high-quality life-reconstruction or even a silhouette can work too, if you can accurately gauge the size of whatever you are scaling it to in it. Permission to use images has already been discussed exhaustively elsewhere.

3) Now you need a program to scale the images with. I’ve seen people do this in paint, or even worse, word. These programs do not, as to my knowledge, offer proper measurement utilities that you will need to make a proper comparison, although I might be wrong about that. In any case there are definitely problems with those programs, as illustrated by the wealth of comparisons with technical problems being posted, and to mitigate those I suggest to use GIMP (or Photoshop, should you be swimming in money that you don’t know how to get rid of). You can download gimp here→, it is free, available for all platforms, should be easy to install for anyone, and its use is fairly self explanatory.

I will walk you through the three simple steps of scaling an animal skeletal:

Firstly, open Gimp and drag and drop the image you want to use into it.


There is a tool called "measurement" (shortcut shift+M) that you can use to measure distances and angles in the image. Here used on the example of femur length (note that in this example the image actually already has a scalebar, and chances are that you should scale to that instead of some skeletal dimension that you might be getting wrong, but this is just for illustrative purposes). The measurement turns up either in a separate window, or in the bottom status bar (as shown here). Note down or keep in mind that measurement, as it tells you how long the thing you are scaling to currently is.


Now use the scaling function. Make sure the chains next to the fields are connected (click on them if they aren’t), as you want the aspect ratio of the image to stay the same. Now divide the height or width by the figure you measured, and multiply by the measurement the structure should have in real life. You can choose whatever scale you want, in this example it is 1px=1mm. However, if that scale ends up being overly small, you might want to consider making it bigger (e.g. 2px=1mm) by multiplying by some factor. The only really important thing, of course, is to keep the scale the same for everything being compared.

So if after importing the image the femur as you measured it is 1011.6 px long, and you want it 243 px long, width/1011.6*243. Make sure to do this in one step, as otherwise you will first scale the image down massively and then enlarge it again, which will result in crappy resolution. Gimp has no problem with you just typing the entire calculation into the scale dialogue, it will handle the rest for you.

Also, I suggest to include a scalebar so it is clear to everyone what the images are scaled to, if there isn’t already one in the image that you used to begin with. This is most easily done by using the rectangular selection (press "R" or click on it in the toolbox), setting the dimensions of that selection to the dimensions you want your scalebar to have (in the tool options dialogue on the bottom left, find the fields labeled "size"), then using the fill tool (shift+B) to fill it with solid black.

Then you can save the file as an image using file>export as. No need to make a screenshot with poor resolution

You can of course perform much more advanced tasks with this software. For example, there is the path tool for creating curved paths, and a plugin→ that returns the length of said paths. Used in combination, these allow you to measure the length of curved lines, which is very useful to measure the axial length of dinosaurs, measured along the neural canal.

There is also the histogram function (under windows>dockable dialogues>histogram), which allows you to measure areas (in px) of the same colour.

I won’t go into more detail as this is about the basics, but if someone after reading this still has problems doing it, I am available to help you.

dinosauria101

[creature386]

There is a free version of Photoshop to my knowledge. It's old and noone uses it, but I like it more than GIMP. No idea if you can take measurements with it though (there's a reason everyone takes the costly version, after all).

[Infinity Blade]

Animal head weaponry: form and function
Quick note: this arose out of AvA debates, but this is still a legitimate aspect of animal functional morphology and behavior, so I believe it warrants its own post for this thread. If anyone reading this can help me turn this into a broader, more encompassing post (I call this "animal head weaponry: form and function" but only talk about adaptations for ramming), it is more than welcome.

Cranial weapons come in a variety of shapes, sizes, and, as it follows, functions. However, an assumption I have seen often made about animals equipped with horns, tusks, and antlers is that they are all somehow automatically adapted to literally ramming their cranial weapons into another, similarly large animal or object at top speed bighorn sheep-style. Imagine a rhino or elephant charing at an animal just as large as itself as fast as it can and colliding the tips of its horn or tusks into the other animal’s body as if it were a horseback knight ramming/impaling enemies with his lance.

The problem with this is that just because an animal has cranial weaponry (like horns, tusks, or antlers) doesn’t mean it is adapted to ram into a similarly large animal. Likewise, there are different degrees to which an animal is adapted to withstand ramming forces to its skull and cranial weapons.

Charging rammers:
Here's an obvious, water-is-wet example of an animal that can charge and ram. The muskox can backpedal, charge for some distance, and collide headlong with another muskox (1:26 to 1:32; subsequent examples shown afterwards).


Bighorn sheep are another example of this kind of rammer.[1]

"Bighorn rams famously collide with their horns by rearing and falling towards each other, which produces loud reverberations. However, they also charge and collide with the tops of their heads between the horns, a direct head-butting behavior similar to head strikes suggested for Stegoceras."

Non-charging rammers:
A potential example of an animal that can ram, but does not appear to engage much in charging and colliding behavior, is the elephant. Elephants are described as sometimes ramming head-on in intraspecific fights.[2] However, my personal communications with elephant ethologist and conservationist Dr. Joyce H. Poole have revealed that, while elephants ram each other close-up to each other, they do not really ram each other like goats, rams, or buffalo do (she told me that one tuskless female did charge and hit her car, BUT (her 'but' was in all caps like here) she did not believe the honeycombed forehead with nasal passages could withstand the kind of force with which the aforementioned charging rammers strike with).

See this video below for what elephant ramming looks like (from 2:11 to 2:16).


^As you can see, the elephant's are quite close to each other from the start. They do not run (or rather, amble; elephants can't run) for some distance and collide heads/tusks like muskoxen do.

Note that this applies to animals as large as the elephant. I won't rule out the idea of an elephant charging and colliding into an animal far smaller than itself with its head and tusks (if it can catch it).

Other examples of this may include giraffes, carnotaurines, and the marine iguana (Amblyrhynchus cristatus). Giraffes strike each other with their ossicones using their long necks, while marine iguanas engage in low-motion headbutting and shoving matches at low speed; carnotaurines are proposed to have engaged in either one of these behaviors.[3]

Non-rammers:
An example of an animal with cranial weaponry but is not suited to any sort of ramming is the mountain goat. The mountain goat's skull is thin, very light, and fragile.[4][5] So instead they use their horns to hook and jab at each other.[5]

Cattle in the genus Bos may be considered another example of these (read further below).

---

Here's a video that I think illustrates all of this quite well:


^The ram in this video collided headlong with a cow (one that was larger than itself, in fact!). But while the ram was able to get back up after the collision, proving that its skull could withstand the impact, the cow does not, at least not immediately afterwards. Evidently, the ram's skull could withstand a clash better than the cow's could, so here we have an example of an animal that can withstand a high-energy, charing collision (the ram) and an animal not adapted to this at all (the cow).

(There is also an upsetting video where two fighting bulls clash headlong, but both die instantly from the impact. It would appear that the genus Bos was not meant to withstand these sorts of impacts)

References:

[1] Snively, E., & Theodor, J. M. (2011). Common functional correlates of head-strike behavior in the pachycephalosaur Stegoceras validum (Ornithischia, Dinosauria) and combative artiodactyls. PLoS One, 6(6), e21422.
[2] Estes, R. (1991). The behavior guide to African mammals (Vol. 64). Berkeley: University of California Press. p. 263, 265.
[3] Delcourt, R. (2018). Ceratosaur palaeobiology: new insights on evolution and ecology of the southern rulers. Scientific reports, 8(1), 9730.
[4] Geist, V. (1966). The evolution of horn-like organs. Behaviour, 175-214.
[5] Yale, K. (2008). Mountain Goats of Glacier National Park. Farcountry Press. p. 24.

[dinosauria101]

theropod

Poor resolution? How? All you have to do is click on it and you'll get good resolution.

With regards to screenshots vs images, I find 3 primary advantages of screenshots:
-They never 'break' the way images do
-They're safe from copyright
-They are faster to make, not to mention easy and free

With regards to the scaling, I was following the 10 percent dimensional increase=33% mass increase rule - I took the upper bound Baryonyx estimate (and the one Hartman seems to have scaled the skeletal to) and scaled up GSP's estimate to get mass for it. And a 10-11 meter Suchomimus is up to 5.2 tons, no? Wouldn't 5 plus tons work for the ~12 meter adult estimate?

[creature386]

Oct 28, 2019 20:33:37 GMT 5 dinosauria101 said:

With regards to screenshots vs images, I find 3 primary advantages of screenshots:
-They never 'break' the way images do
-They're safe from copyright
-They are faster to make, not to mention easy and free

He means images made by exporting a size comparison as an image file rather than screenshotting it.

How are screenshots save from copyright? Screenshotting a copyrighted image and then uploading it is no different from sharing it otherwise.

[dinosauria101]

Oct 28, 2019 20:45:55 GMT 5 creature386 said:

Oct 28, 2019 20:33:37 GMT 5 dinosauria101 said:

With regards to screenshots vs images, I find 3 primary advantages of screenshots:
-They never 'break' the way images do
-They're safe from copyright
-They are faster to make, not to mention easy and free

He means images made by exporting a size comparison as an image file rather than screenshotting it.

How are screenshots save from copyright? Screenshotting a copyrighted image and then uploading it is no different from sharing it otherwise.

What I meant by that was that my screenshots are effectively safe from everyone we DON'T want using them. Specifically Taipan and his pseudo-minions.
Guests cannot view screenshots and when the URL of a screenshot is copied it gives you a very tiny and blurry image. Moreover, I bet you a good majority of my screenshotted size comparisons are naturally unappealing to them because of what they compare.

I've made accounts on Photobucket, Imgur, and ibb and have had far too much trouble with uploading images. Screenshots work much better for me.

[creature386]

theropod was not suggesting anyone to use Imgur or Photobucket (though I can explain you the former, if necessary).
You can still upload the attachments. Theropod was merely suggesting to export the images with the program they were made in rather than screenshotting them.

[dinosauria101]

Those were merely reference points for uploading stuff. My point was merely that exporting images scaled properly is lengthy and tough; why bother when there are nice and easy screenshots to be made?

[theropod]

Oct 28, 2019 20:33:37 GMT 5 dinosauria101 said:

theropod

Poor resolution? How? All you have to do is click on it and you'll get good resolution.

I know that I can click on attachments to enlarge them, I post attachments all the time. The resolution has nothing to do with whether you upload it as an attachment here, or to some image hosting site, or something else. But making a screenshot is still limiting your resolution
Example:
compare that to the last comparison I made, which I also uploaded as an attachment:


See what I mean?
If you are making a screenshot, maximum resolution of your image equals maximum resolution of your screen. Keeping in mind that you might not be able to make what you are actually trying to show (i.e. the actual comparison) fill your entire screen (as in the example above), that might not be good resolution. But using a real image editor, you can produce images in any resolution you want.

With regards to screenshots vs images, I find 3 primary advantages of screenshots:
-They never 'break' the way images do

Screenshots are images, same thing. You are confusing screenshots with attachments. This is an issue with where you host your image, not whether it is a screenshot or not. You can upload an image here that isn’t a screenshot, I do it all the time, they are the same.

-They're safe from copyright

They are not "safe from copyright" in any way.

-They are faster to make, not to mention easy and free

And I just described to you an even faster, even easier, and even freer (word costs a lot of money, gimp doesn’t) way to make comparisons, that is on top of that more accurate and produces better results. Take it or leave it, I don’t care. I just thought you’d be interested, since you are posting one comparison after another, all of them screenshots, and apparently done in word for some reason.

With regards to the scaling, I was following the 10 percent dimensional increase=33% mass increase rule

I don’t know what you are talking about here. Assuming this is in reference to this post, your scaling is off again: theworldofanimals.proboards.com/post/48867/thread

Please take another look at the post I made on this thread, specifically about proper scaling. You need to cube the size ratio and multiply by the mass.

If a 9 m Baryonyx is 2.5 t, then an isometrically scaled 10 m Baryonyx is 3.4 t, and a 12 m one would be 5.9 t. But the premise is wrong, as the first two are the same specimen, so it is likely that one reconstruction is simply longer, not necessarily much larger overall. In order for isometric scaling to make sense, it needs to be a case where we can actually expect isometry. If it’s the same specimen, but one guy restores its snout and tail longer, or something like that, then by definition we do not have isometry.

I took the upper bound Baryonyx estimate (and the one Hartman seems to have scaled the skeletal to)

That is not how skeletals are made. He didn’t "scale his skeletal to that estimate", he reconstructed the skeleton bottom up, and based on the sizes of all the available bones the skeletal ended up being that long.

[dinosauria101]

theropod

Okay, regarding resolution, I am able to post in higher res given the chance; I often do so on WildFact and can do it here too
Attachments vs screenshots: my bad. Attachments never break
Copyright: Well, they may not be SAFE, but they are both hard to post well and probably unappealing to Greg Rhodes and his minions like I said above
Word vs other programs: Well, I appreciate the suggestion. But Word is going smoothly so far like I said.
Scaling: I will have to check Paul's skeletal when I get the chance
Dimensions: that is what it looked to be based on the scalebar matched with a size chart.

[creature386]

As theropod explained scaling, how about a list of useful programs for a paleontologist/biologist?

I will not write any tutorials, as that would take way too long, so let me just give you a list.

-Gimp:
www.gimp.org/
It's uses in creating size comparisons, scaling and taking specific measurements have been detailed above, so I don't need to go over that. You can also do Photoshop-like image manipulation with it, if you feel like that is necessary. Free.
Photoshop can be used for many of the same goals, though the good versions aren't free.

-Past3:
folk.uio.no/ohammer/past/
A rather obscure program that has been recommended to me by at least two paleontologists. It's icon is an ammonite, can it get any more obvious? It is useful for plotting data and doing statistical analysis. Free.

-R:
www.r-project.org
Okay, this is more of a programming language than an actual program, but I'm adding it for the sake of completion. R is used for creating graphs, tables, and statistical computing. Widely used in science and free (I've heard there are places where you have to learn it if you want to study paleontology, but I'm unsure).

-Python:
www.python.org/downloads/
Python is a programming language that can do most R can do and it's free, too. Personally, I find R more suitable for statistical analysis and its graphs are prettier, but Python is, IMO, better at handling large datasets and creating functions to solve problems.

-BioEdit:
bioedit.software.informer.com/7.2/
BioEdit is a sequence alignment editor and if you want to get into genetics, there is almost no way around it. It's good for manually editing DNA sequences, regardless of whether they have been aligned or not, and at producing useful chromatograms where you can look at the peaks in a sequence's nucleotide distribution.

-Mesquite:
www.mesquiteproject.org/
What BioEdit is for genetics, Mesquite is for taxonomy. This is a program in which an evolutionary biologist or taxonomist can store tables of organisms and their diagnostic characters for further cladistic analysis.

-TNT:
www.lillo.org.ar/phylogeny/tnt/
TNT is often used in synergy with Mesquite. Mesquite produces useful taxon-character tables which TNT can then turn into phylogenetic trees and do bootstrap analysis on them.

-Tilia:
www.tiliait.com/download/
What BioEdit is for genetics and what Mesquite and TNT are for evolutionary biology/taxonomy, Tilia is to palynology. This software is excellent for creating tables of pollen taxa and their frequency and turning them into informative diagrams.

-Excel:
OK, technically, this is less of a scientific program, but more of a program absolutely everyone uses. Good for noting down your measurements in tables which you then analyze in Past3 or R. You can also make graphs' with it, but they are uglier than R's and Python's. You can use it for free here, though a desktop version costs money, unless you want one of the tooth-less versions floating around (it's similar to Photoshop in this regard):
office.live.com/start/Excel.aspx

-Inkscape:
inkscape.org/
If you have a photo or a drawing of a fossil and scanned it into your computer and now you want to turn it into a vector graphics image to use in your latest paper, use Inkscape. Free.
CorelDraw can achieve many of the same goals and it's free, too (I don't really know the differences between the two programs' capacities).

[theropod]

Erasing stuff you don’t want in your images:


1) Open your image in Gimp (right click on image file>open with>select "Gimp" or start Gimp>file>open>select file)


2) Hit shift+E or select the eraser from the toolbox and set it to the desired size in the tool options below.


3) Erase whatever you want to erase.



optional 4) You can also use the rectangular selection for removing pieces of the image. For that, hit R or select the rectangular selection in the toolbox.


optional 5) Then drag the rectangle around what you want to remove.


optional 6) Now, hit "delete" on your keyboard. Now save the image (File>export as). Finished.

[creature386]

Don't forget the magic wand! It's extremely useful to get rid of homogenous backgrounds.

[theropod]

Correct. However that is not strictly necessary just to make a basic comparison. There are lots of good gimp tutorials around, and to give one would be beyond the scope of this forum.

[creature386]

That depends on if you are one of those people who likes to get the animals into the same background, even if they were originally from different ones. There are a few people who like doing this, though it's an issue for most posters, as we like making comparisons with animals from the same white or squared backgrounds.