Topics

61

General / Intelligent scissors are not Intelligent

« on: June 24, 2019, 08:36:17 PM »

Hi,
I am following the pdf on use of tie points so as to mask all photos when having masked one or two using as it suggests intelligent scissors.

There are no videos on you tube of intelligent scissors in use.
I search in the program's help on intelligent scissors but it finds nothing.

I use the tool but it makes straight lines around my curved edge, I expect it to snap to the edge, in an intelligent way.
I go Ctrl  A but have a selection still not following the curved edge.

What is supposed to happen as all they give me is straight line cuts.

Steve

62

General / can I have both 1.2.6 and 1.4.5 installed to use both ?

« on: June 24, 2019, 02:50:23 PM »

Hi,
my exhaust stack was far better in 1.2.6 but I need access to the suggested masking method 'apply masks to tie points' available in 1.4.5

Can I have both versions installed on my PC and use whichever suits me for what I am doing ?

Does installing 1.4.5 remove 1.2.6 ?

would that happen if I installed to D drive ?

Unless I hear back on why 1.4.5 gave a poor result and metashape a complete mess on the same set of photos I need to still use 1.2.6

would projects in 1.4.5 be backwards compatible, i.e. after using the 'apply masks to tie points' in 1.4.5 could I process them in 1.2.6 ?

Steve

63

General / align chunks then show aligned chunks sees the rotate object ball icon disabled

« on: June 24, 2019, 02:41:50 PM »

Hi,
1.2.6
Having run workflow align chunks camera mode, then click on show aligned chunks icon, there is no tool for rotate object, its greyed out.
in the video here at 2m 20sec her rotate object is not greyed out yet she has turned on show aligned chunks.

https://www.youtube.com/watch?v=pYE0KOa5oGk

is this something that is only allowed in Pro ?

Steve

64

General / localised decimation option would be FANTASTIC ?

« on: June 24, 2019, 02:28:21 PM »

Hi,
 if one types in 200,000 for mesh decimation, that could be good for an egg cup but bad for a car.
we are saying its got too many faces for what is needed for accuracy, but I have been using a value seen on the internet of 200,000 which I am now thinking is that making my model a bit crude if its 50cm long, and very crude if its a car. If its very ornate is that bad, if its very simple with little detail is that still too much ?
should it not simply be we take 1/10 of the current declared face count and enter that ?

Are we left to decide something that the program would be better at and know best value on given the complexity of the surface ?

If I use 400,000 as I have to see difference on a 38000000 mesh and I am trimming away the base grid I find some pink triangles bite into the model so in fact I need more to avoid that.

advice please on best approach.

Steve

65

General / How does one draw straight line selections ?

« on: June 24, 2019, 02:19:50 PM »

Hi,
 I have a lot of straight edges to trim points and mesh from, I try to use shift and click with the lasso tool to get a straight line but nothing happens.
Obviously the circle tool isnt of use, but a variant on the lasso with straight lines I truly expected in a prog of this calibre.
I cannot draw straight lines with the lasso though I am very good compared to other average mouse users.
Just where is this basic tool we need ?

I am with Standard but hoped for something  more than plastic scissors !

I am beyond teething stage and nappies.

Struggling here with lack of basics such as scale, merge flags and now straight line selection for straight edges.

Steve

66

General / Why spend time generating 38million mesh face to then decimate it ?

« on: June 23, 2019, 07:41:06 PM »

why is it necessary to end up with mesh count 38,000,000 to then have to decimate it to 200,000 ?
It takes time to create 38million faces then we are told to decimate it.
I find the whole model is sluggish until decimation is done.

this time I select 400,000 not 200,000 incase 200,000 was too simple.

if 200,000 is ok  why are we wasting time on 38,000,000 ?

Steve

67

General / Basic tool to assist bringing models together not in basic program anymore.

« on: June 23, 2019, 06:37:40 PM »

Hi,
I had to turn the object upside down to photo its base, and such will be something many other users will need to do.
This is just basic workflow for those requiring the basic simple program, objects do have bottoms !
This does not require Pro at 3500 price versus 179, a price that is to pay for its  GIS geomapping and other advanced stuff which I and others have no need of whatsoever.
I come to add markers as per the video on youtube bringing together a flint . https://www.youtube.com/watch?v=pYE0KOa5oGk

I simply need for basic work markers to tie the two halves of the flint together, not interested at all in georeferincing or other such advanced things, but just the basic task of telling agisoft how to bring two models together ignoring the baseboard that is obviously otherwise going to confuse the hell out of the program as it has moved relative to the object when object turned upside down.
I have deleted it out in mesh of each chunk keeping in a portion for scale.
I also see scale only exists in pro, again thats a basic need to give a size to the model, must we pay 3500 v 179 just to have a size for it ? Surely standard use should see a size being possible ?

There are no markers as such anymore in the standard edition. This basic tool has been removed. Other users have also posted on this loss, they used to have access to markers so as to help bring two meshes together, now removed.

https://www.agisoft.com/forum/index.php?topic=2524.0


Standard edition as such cannot do a standard job anymore.

As others have also said they dont need all the advanced extras and a price tag of 3500 to pay for them.

It should be though that a basic tool to assist bringing two models together for what is a standard requirement exists in the basic program.

What other method then is there to perform this standard requirement ? I have bought Agisoft to do this, and after 5 days intensive photography with 5000 photos  I have lots of objects I have photographed where I have had to turn them upside down. Now the marker tool is no longer there.

Its basic model making from objects on the floor or a gridded table. not for profit, I am not a business, its a hobby. I dont have 350 let alone 3500.

Pro has now the markers that used to exist in standard, as posters say. 3500 versus 179  I cannot find that money just to assist bring two models together making Agisoft ignore the baseboard. A standard function which should be in standard edition. merge chunks by points sees it use baseboard and I get a mess.

I am gutted. just dont know what to do now. 179 spent and stuck, basic wont do basic.. and only the old version will work on perfectly ok photos the newer versions fail on.

Steve

68

General / What is the method if I have a second set of photos of underside of object ?

« on: June 22, 2019, 07:52:14 PM »

Hi,
What is the correct procedure when an object needs turning upside down ?

I photographed around the object with it the right way up.
I turned it upside down to photo it again showing its underside.
I have put the photos with table and floor (unmasked) into the program , I get two objects, I also see the table and floor, I can delete the unwanted table and floor out in the mesh or point cloud.

Does one create a project per set of photos or a chunk per set of photos ?

I took the photos of the underside in chunk1 and made a chunk2 with them, then redid align photos on chunk 1, and got a single object. I then selected chunk 2 highlighted it and ran align photos but still see point cloud of the object upright, how do I get to see the object for chunk 2, I click on chunk 1 or chunk 2 but the object in 3D view remains same.

I expected it to show the object from chunk2 pics if chunk2 was clicked on.

If I then merge chunks does the program bring the object together adding the underside mesh to the rest of it or do I end up with 2x the mesh I had before ?

I fear that if I merge chunks, and Agisoft matches the table top and floor, I will end up with two objects with a common floor again.

I need to have Agisoft make the two models, then I delete the points for the floor from each, then bring the two models together. I havent the time to go masking lots of photos, I have 5000 photos of a whole load of objects and its far quicker to delete the points !

How is this bringing of two models together done after I have deleted out the unwanted surroundings ?

Steve

69

General / Should unwanted items such as plants surrounding the subject be edited out ?

« on: June 20, 2019, 05:54:41 PM »

Hi,
Can photoscan cope with background textures unwanted in the final 3D object, such as vegetation or gravel etc ?
I photographed a vintage vehicle open beam chassis after clearing away plants that had grown around it but I only flattened them. Also photographed items placed on someones tarmac driveway, others on a concrete floor, another in a museum, another on bare soil.  To have to edit white in up to the edge of the object in every photo, of which there are hundreds, in fact a few thousand, will take forever. Can photoscan deal with plants etc or the fact that they move whilst the object is static, will that cause failure or inaccuracy ? When the item gets turned on its side or upside down to photo its underside, how does photoscan cope with the changes ? What about the items in the museum on the museum floor or on the walls or structure in the ceiling ?
Should all this be edited out as the mask method before the photos are put into a new photoscan project ? many hundreds of photos, thousands in fact to do if that is required.
Its not always possible to be in a nice studio with the object on a white background with flash umbrellas !
Can I not remove the resulting unwanted areas after the point cloud is made in the first stage.
Worried that Agisoft will find this extra distracting 'material' a problem and fail in processing.

If it has to be masked out, must it be accurate pixel perfect right up to the edge or will a rough trim leaving a bit showing around the object which would be far quicker to do, be enough ?
Steve

70

General / no operating systems listed for metashape

« on: April 20, 2019, 07:21:21 PM »

Hi,
there is no indication anywhere on the requirements page of what operating system metashape will work with. Installing a program on a pc that doesnt have the correct operating system can cause problems, Why must users risk that happening to their pc to find out ?

Should Metashape work on windows 7 64bit pro ?

I just cannot get it to launch, I get a black filled floating cmd window and pc locks up.

Steve

71

Bug Reports / metashape trial wont launch and freezes up PC win 7 64bit pro

« on: April 20, 2019, 07:07:23 PM »

Hi,  I have just installed Metashape 64bit demo , I was running Photoscan perfectly ok yesterday. I create shortcut to the exe file, double click it, get a black dos type window, task manager unable to end task metashape, I am even unable to close down PC, I eventually get PC shutting down, but it wont after 5 mins shut down, I have to hold power button in, if I was running raid this would lose all synchronisation and mean a total PC rebuild. This prog needs testing on a win7 64bit pro pc with anything that might mess it up, inc norton, kaspersky and others.  What a cock up !
I have spent ages creating photo sets to test this on and its messed up the pc instead. Photoscan never did this.

Also there is nothing on the agisoft website to say what operating systems this works with, and nothing to warn of potential problems and what to do to get it to run.

I have added agisoft folder to excluded list Kaspersky. then run metashape and yet again we get the black dos box, which I cant X close, PC is locked up.
I have had to reinstate the acronis backup made just prior, glad I did do that. I shouldnt be having to do a restore after installing Agisoft.

This is terrible, so disappointing, had hoped to buy metashape.

will this metashape in theory run on win7 64bit pro ?

How can I get this to run, someone mentioned that the beta version solved this for them, where is that beta version ?

Steve

72

Face and Body Scanning / Will Agisoft work with 1 camera if human stay dead still and hold breath ?

« on: September 13, 2016, 01:41:25 AM »

Hi,
I need to create a 3D model of a human male wearing nothing but just shorts and socks rolled down over shoes, kneeling down and another separate session with same guy standing up holding arm up, it will be far easier to photo someone than try and create this in 3D programs, especially trying to model the clothing, hours versus days of work. If I place three garden canes tied together at top as a triangle  to support the wrist or even use my photo tripod to support his elbow and move around taking photos very quickly on a canon 70D in RAW mode fast shutter speed 1/500 f8 800asa  (freeze movement, good DoF and lens sweet spot ! correct exposure cloudy bright), on manual exposure on a cloudy bright day with no breeze, so no shadows or material movement, is there a good chance of success ? If a part of the human doesnt move relative to another part of the human, surely it will work ?

If I give it a few sessions then knit the parts that do work together and heal up the gaps maybe ?

What prog/software is there that is affordable on a tight budget or free to heal the gaps up if I need to do it this way ?

Has anyone tried such and is it just impossible ? How fussy is the prog if something moves a few mm on something human size ?

I cant see there being a problem if human sways a bit as thats no different to me moving to and fro as I shoot.

What about the bushes or grass though, they move, do I have to paint them out of the picture in photoshop ?

What about things in the background of a photoshoot in fact ?

Will it be ok on areas of skin if photographed on a current dslr and a very good quality lens ?

Steve003