Topics

1

General / Stumped. What is going on here?

« on: September 12, 2016, 01:32:03 AM »

I'm shooting a small, yellow jewelry case.  It has a faux lizard texture on it.  While the box is symmetrical, no one face is the same as another in terms of texture.  In other words, a lizard skin is very textured.  Please see screen shots attached.  There is lots of variety.

As you will see, I have taped over the very reflective handle on top.  In Photoscan, I have also masked out the shiny clip on the front.

Jeez - this thing WILL NOT ALIGN!

I've tried placing points.  Photoscan cannot recognize them for all photos.  For example, I placed a marker on the tip of the clasp on the front.  Photoscan lined it up pretty well in the first few shots, giving me a small grey icon, and for the rest it put it in the wrong place.

When I "Estimate Image Quality" some shots that are tack-sharp are reported as below 0.5 even though it's my understanding that this number is a reflection of the most in-focus area of the image.  Even when I disable these below-0.5 shots, still the box will not align.

I would think that despite the symmetry of this box, Photoscan would be able to identify unique points on the lizard skin finish and work it all out.  But apparently not. So why is this?

I'm shooting 4K images, on a tripod, ISO 100, f14 with a 2-second shutter on a 70mm zoom lens with a cable release to reduce/eliminate camera shake.  I am capturing RAW and processing in Photoshop to 16-bit TIFF files.  I am masking every image in Photoscan.

I've included 4 sample low-res shots of the box itself.  These are the shots I am feeding into Photoscan.

I've also included a screen shot of my photos pane and my model window so that you can see the results I'm getting.

I'm not new at this.  I've done hundreds of models, including people and never had this problem.  I picked the yellow jewelry case because of its highly irregular texture, thinking I'd make a model out of it in a couple of hours, taking it slow.

2 days later and empty-handed I'm posting this.

Thanks in advance,

Jon

2

Feature Requests / Masking by focus

« on: September 10, 2016, 09:53:29 AM »

It would be super-useful to me to be able to automatically mask an image by sharpness of focus.

I shot a small horse model today.  It's like a kid's toy.  I shot it in parts, meaning, in one shot I'm focusing on the neck, in the next shot I'm focusing on the tail.  But my frame is taking in more than the exact thing I'm focusing on and so some areas of the horse are soft-focus and some are downright out-of-focus.

I was thinking that it would be great to be able to tell Photoscan to mask out the out-of-focus areas and to be able to determine, by using a slider similar to the gradual selection tools, what Photoscan determines as out-of-focus and what it determines as in-focus.  That way you could dial it in.

I think that Photoscan does some filtering automatically (in the background) of out-of-focus areas but I don't know that for a fact.

But as Photoscan generates (or at least calculates) a depth map, it must know what is in-focus and what is out-of-focus already.

Just an idea.

Jon

3

General / Too many photos?

« on: September 02, 2016, 01:50:03 AM »

In Photoscan, if you don't take enough photos the program has a hard time working it out and making a model.

Is it possible to take too many photos?

And if so, where do you draw the line?  Meaning, where does it become too much, if indeed you can take too many pix for the program to be comfortable with?

Jon

4

General / What Photoscan is doing

« on: August 19, 2016, 09:07:38 AM »

Is Photoscan locating contrast or colours as well in the photos that you feed it?

I have a plain, lightly textured box that I want to make a model from.  If I put small 1/8" dots all over it, all different colours, is Photoscan going to pick up on the colour difference and create my box or will it see all the dots as black and white and just recognize the contrast?

Without any means of identification, this box will not come out at all.

I get it that as long as these dots are randomly placed, it should work it out.  But still, I'm curious.

Thanks.

Jon

5

General / Alignment question

« on: August 12, 2016, 09:03:24 AM »

I shot a baseball bat and because I wanted a super-detailed texture, I took a lot of close-up shots with a 200mm lens.  I made sure that I had very good overlap between shots and I started at one end of the  bat and worked my way to the other.  The bat was positioned vertically on a turntable and I turned at approximately 20 degree increments between each shot.

In Photoscan, I got a terrific sparse cloud when I aligned my photos on "Low" and a pretty poor sparse cloud when I set my alignment to "High". In fact, on "High" only half of the bat showed up.

Why is this?

The photos were all in-focus and above 0.5 on the quality scale.  The camera was on a tripod and everything was correctly exposed.

So why did Photoscan not align in "High" mode?

Jon

6

General / Use of markers

« on: August 08, 2016, 05:28:32 AM »

Can markers be placed manually on photos to assist in alignment?

Let's say you're shooting something that has minimal recognizable qualities and those that it does have tend to repeat (like a basketball), making the object symmetrical.

Can I, as the user, place markers on photos that identify definite points on the object?  And then "align photos".

Or not?

Thanks.

Jon

PS.  If not, it might be handy to implement this.

7

Feature Requests / Lasso tool

« on: July 31, 2016, 07:29:49 AM »

A lasso tool in addition to the existing selection tools would be terrific.

Jon

8

General / Aligning chunks

« on: July 30, 2016, 09:15:28 AM »

Can I align chunks by placing markers on the original photos?

I shot a rock (woo-hoo).  I have 2 sets of photos: one with the rock facing down in a lump of putty and the other with the rock facing up in the same lump of putty.  Each batch of photos is in its own chunk.

I obviously want a whole rock, not two parts.  So I need to align these chunks. 

Can I do so by placing markers on the original photos or do I need to carry each chunk through to 3D model stage and then place markers and align?

Jon

9

General / Automatic distance calculation

« on: July 17, 2016, 07:42:36 AM »

This might seem like a stupid question but:

If I have known distances between markers in my project, can Photoscan calculate distances between markers for which I do NOT have a known dimension.

Example:  I shot a room with a desk and a window.  I know the exact dimensions of both desk and window.  I place markers accordingly, telling Photoscan coordinates and distances that I have measured in the field on both the desk and the window.  So I have 4 markers on my desk and 3 on my window.  Photoscan tells me that my error once all measurements are input is 0.018656 meters.  Fine.  I can live with that.  I'm not doing rocket science.

But I want to know how far it is from one of the markers on my desk to one of my markers on the window.  So I select both markers in the Workspace menu and create a scale bar.  The scale bar shows up but of course, there's no distance entered because I don't know the distance. 

Can Photoscan calculate that distance for me?

And if so, how?

And if not, wouldn't that be a useful feature to have?

I'm positive that it can.  I just don't know how to do it.

Thanks in advance.

Jon

10

General / EXIF data

« on: July 15, 2016, 08:47:17 AM »

Photoscan will not process images with no EXIF data, correct?

A friend gave me a .mov file of a building that he shot with a drone, circling around it.  I converted the .mov to .tif files and imported them into Photoscan.  I had a feeling that they wouldn't work. And they didn't.  They didn't have any EXIF data.  It was all stripped out in the conversion.

Jon

11

General / Photoscan Markers

« on: July 11, 2016, 03:31:32 AM »

In shooting a limited area, approximately 150 meters x 150 meters, where I am interested in the ground plane and what's on it (as opposed to trees and other elements that project up from the ground):

a.  Does it matter where I place my Photoscan markers as long as they are visible to multiple cameras?

b.  Does it matter how many Photoscan markers I use? 

c.  Does it matter which Photoscan markers I use? There are a lot to choose from.

d.  With Photoscan markers placed, do I need any other measuring device to appear in my shots, such as a 3 meter long stick or anything like that?  Or does Photoscan automatically recognize its own markers and also know how far apart the markings on them are and thus scale the scene correctly?

By "Photoscan markers" we are obviously talking about the black and white ones that you can print from Photoscan.

And finally, I've read in other posts that there is a command in Photoscan to detect markers in photos automatically.  Where exactly is that command located?

Any input would be appreciated.

Jon

12

General / Photoscan markers

« on: June 29, 2016, 05:04:46 AM »

Okay, so I found out all about Photoscan markers.  They're great.  Wish I'd known about them earlier.  I think we should mark it up as something to put in the next revision of the manual.

Now:  Is there a recommended USE for these things?  Should I put one entire page of 5 markers on the floor next to my model?  Should there be one or more placed on boards that are (for example) 3' above the floor?  Am I supposed to cut them out and use them individually?  And finally, is there any further significance to these black and white circular shapes that I should know, such as "from the white center dot to the outside edge is exactly 26.753mm"?

Maybe someone can point me to a post in the forum or an external link that answers all of (or some of) my questions.

Thanks.

Jon

13

Face and Body Scanning / Nylons

« on: June 28, 2016, 08:16:13 AM »

I'm having a hell of a time getting a decent model from shots of a woman wearing nylons.

My photos are super-sharp and I'm guessing that my problem is with the moire that these nylons create.

But I'm posting on the off-chance that someone has already experienced this, knows why it's happening and how to handle.

To be specific, the model is extremely bumpy.  And I mean really bumpy.

Thanks.

Jon

14

General / Textures

« on: June 23, 2016, 07:35:44 AM »

I exported a texture for a model at 4096 x 1.  Looks great.

To see if I could improve it when it renders in VRay, I exported it again at 8192 x2.

Hmmmm...

Why am I getting two (2) TIFF files now?  I guess it sounds sensible, seeing as I checked "x2".  But I don't understand what this is all about.  Nor how to use these 2 files.

There is a reason that "Newbie" is still next to my user name on this forum.

One day I'll graduate to at least a Junior Member.

But not until I understand things like this texture business...   

Thanks in advance.

Jon

15

General / "Camera Stations"

« on: June 23, 2016, 01:50:40 AM »

I shot a china doll under perfect lighting conditions (strobe) and from a number of static camera positions.  The first position was head-on to the model, the next slightly angled down, the next slightly angled up, etc.  I did this to ensure that I captured all undercuts in the doll and the top of her head.

The doll herself was standing on a circular scale sheet with various markings so that I can indicate scale in Photoscan.

I rotated the doll by hand; she was not on a turntable and so her exact center may have varied somewhat but this would have been very minimal.

I took about 15-20 shots for each revolution.  The shots were taken at ISO 200, 1/125, f11 and are nice and crisp.

It's my understanding that these photos need to be separated into different camera groups within the same chunk once imported into Photoscan and that these individual camera groups need to be named "Camera Station."  I read this in the manual.

So far so good I think.

A few questions:

1.  Is this part of the workflow accurate?  In other words, does anyone see anything wrong with my sequence?

2.  Can I name these camera groups "Camera Station #1", "Camera Station #2", etc.  Or do they all have to be called just "Camera Station"?

I'm posting this because I'm not getting the best results.  The dense cloud is coming out flat and with no depth which mystifies me because this is a project being done under controlled and perfect conditions.  But my results are way, way better when I go out in the field under a cloudy sky and shoot a wall hand-held (for example).

Thanks in advance.

Jon