XREF, insert, MAPIINSERT an ortho-rectified aerial image into AutoCAD/Civil 3d

[pbradburyYWE5N]

First let me say I am a long time Microstation user and recent convert to AutoCAD, specifically Civil 3D. As you might gather the transformation has been aggravating. 

I am attempting to accomplish something that was only a few clicks to do in Microstation and I've consulted every professional in my organization and done weeks of research online to try to do this same process in AutoCAD and I haven't found a solution. Ultimately, I know the first question most anyone will ask, is "send me your files and I'll look into it", if it requires that, then I might assume AutoCAD simply can't do what I'm asking it to. I see this as an issue similar to if I tell an auto mechanic that I'm out of gas and ask why my car doesn't start, the mechanic will likely not ask to see my car... Ultimately, I appreciate any help anyone can provide, but I'd prefer not to receive random suggestions and offers of assistance that don't even address my issue, just to provide input, or otherwise ask me to jump through 100000000000000000000000 hoops while providing no real plausible explanation. 

So on to my issue... With Microstation, if I wanted to download an aerial photograph to use with a transportation project that has accompanying topographical survey, and wanted that aerial photograph to match up, near perfectly with the topographic survey, I could simply download an aerial from ArcGIS, ensure the coordinate system in ArcGIS was the same as my Microstation file and viola it would automatically reference in the exact correct location, without any twisting, warping or moving of the sometimes hundreds of aerial tiles I have to download. 

I have attempted this same approach with AutoCAD, but every single attempt I've made, with changing every conceivable setting in ArcGIS and AutoCAD (including similar settings noted above) the aerial always comes out shifted approximately 12 feet off in one direction or the other.. I have verified there is nothing wrong with the topographic survey by going to the project site and checking the northings and eastings of several points within the survey match up perfectly and have checked against multiple different projects. I have also exported the survey to Microstation and referenced in an aerial in the same fashion noted above and it overlaps near perfectly <0.5 feet off. Can anyone run me through a hopefully straight forward and hopefully simple process of simply getting an aerial photograph to appear in the right location, without wasting hundreds of hours of work on every project and proposal we prepare by having to manually move hundreds of image tiles so the road is actually in the correct location? 

Again, thanks for any help!

[chriscowgill7373]

Perhaps the topo is what is off or the orthoimagery is off in this area.  It is hard to tell, as the image quality is very poor for both the tif and Bing. I brought in the image, and it lines up perfectly with the bing imagery from GEOMAP, and neither one line up in particular areas of the drawing.  There is no International feet in Cali, they only use US foot.

[chriscowgill7373]

Google looks to be about the same as Bing and the tif provided:

[rl_jackson]

Why be silly and change that like the laundry.........

Very true that you make that statement, however it does not have any bearing on the use of C3D....

[Mike.M.Carlson]

Move on @rl_jackson. You have provided no help here and clearly have no clue about current versions of Civil 3D.  Enjoy Civil 3D 2013.

[pbradburyYWE5N]

See attached marked up clip of the particular in-accurate areas between the topo and the aerial. The roadway is shifted over a full lane width.. The centerline is where the ETW should be and the aerial is showing a roadway where the topo shows dirt.. It doesn't appear it is the aerial I downloaded if you folks all downloaded Bing maps and a variety of other map options and it still wasn't accurate... I also don't think it is the topo file being bad. I can post another project topo file and aerial that has the same/similar inaccuracies between the topo and the aerial I can download from ArcGIS or via Bing maps. Mike, you appeared to attribute a particular coordinate system to the aerial from Map 3D is that correct? Does this re-project the image, or does this simply tell AutoCAD what coordinate system it should be? Is this the best accuracy I can expect using aerials with AutoCAD/Civil 3D is 12 feet off where it should be? Can anyone provide an example where they have downloaded aerials before where topo is correct with the aerial and that none of the aerials required twisting or warping to get them to fit? I don't think Bing maps or anything associated with AutoCAD 360 will work as I'd like to be able to create good exhibits that can print to large strip plots where the aerial is tiled and looks presentable instead of a fuzzy mess that doesn't contribute anything to the project understanding and unless I have a bunch of aerial tiles it seems as though the Bing maps option zooms out to fit the printable area of your exhibit? I don't understand why this is so dang hard...thanks for any continued help..

[chriscowgill7373]

with that many sources showing the same thing, it could be that the aerial in that particular area is just wrong.  Perhaps that area was shot from a far side view and not straight on, and that is causing the issue.  If you go up to the bridge on the side street, it appears to only be 6' or so off

[Pointdump]

P,

I'm thoroughly enjoying your thread. This is good stuff.

I created a point on your Tollhouse Road Alignment at 13+75. Then in Excel I multiplied the Northing and Easting by 0.000002 (2ppm), which is a good way to visualize any discrepancy between US Feet and International Feet:
N Shift: ±4.49
E Shift: ±12.88

I moved your image 12.88,4.49 and it looks a bit closer, at that point, to matching up with your linework, but Bing Live Maps remains where it was.

Another consideration is Datum. California is moving. So from NAD83 to NSRS2011 there might be considerable difference. That makes Metadata (data about data) critical.

But at the end of the day, a map needs to be produced. We have the tools we are provided. If AutoCAD C3D isn't cutting it for you, then you probably should ask the Boss if you can use MicroStation, which works flawlessly every time. (insert humorous smiley)

Dave

[Pointdump]

P,

The new <<<Beta Transformation Tool>>> shows the transformation from NAD83(1986) to NAD83(2011) of that point at 13+75 is less than a foot.

NAD83 to NSRS11
N Shift: -0.89'
E Shift: 0.85'

Dave

[Mike.M.Carlson]

Hi @pbradburyYWE5N

In response to your questions: 

Mike, you appeared to attribute a particular coordinate system to the aerial from Map 3D is that correct? Does this re-project the image, or does this simply tell AutoCAD what coordinate system it should be? 

Yes I noticed the dwg file had NAD83 HARN CA SPC Zone IV specified, but in your first post, I thought you mentioned that the Aerial was based on NAD83 (1986) CA SPC Zone IV.  I just switched to make sure everything was apples to apples.  However, based on @Pointdump helpful recent posts, the shift between NAD83 realizations seem small.  

Can anyone provide an example where they have downloaded aerials before where topo is correct with the aerial and that none of the aerials required twisting or warping to get them to fit?

There are many "free" sources out there, but unless you pay for a true photogrammetric or other type of Arial Survey, you are at the mercy of whatever tolerance the source specified for the aerial data.  Without metadata, it will be hard to diagnose.  @chriscowgill7373 has done great leg work to check multiple sources and it seems strange they don't match up with your data, but that doesn't mean your survey data is off... to be honest, a ~10-15 feet shift isn't that bad for free imagery.  I guess it depends on what you're trying to do with this imagery.  

[Cadguru42]

---

@pbradburyYWE5N wrote:

First let me say I am a long time Microstation user and recent convert to AutoCAD, specifically Civil 3D. As you might gather the transformation has been aggravating. 

I am attempting to accomplish something that was only a few clicks to do in Microstation and I've consulted every professional in my organization and done weeks of research online to try to do this same process in AutoCAD and I haven't found a solution. Ultimately, I know the first question most anyone will ask, is "send me your files and I'll look into it", if it requires that, then I might assume AutoCAD simply can't do what I'm asking it to. I see this as an issue similar to if I tell an auto mechanic that I'm out of gas and ask why my car doesn't start, the mechanic will likely not ask to see my car... Ultimately, I appreciate any help anyone can provide, but I'd prefer not to receive random suggestions and offers of assistance that don't even address my issue, just to provide input, or otherwise ask me to jump through 100000000000000000000000 hoops while providing no real plausible explanation. 

So on to my issue... With Microstation, if I wanted to download an aerial photograph to use with a transportation project that has accompanying topographical survey, and wanted that aerial photograph to match up, near perfectly with the topographic survey, I could simply download an aerial from ArcGIS, ensure the coordinate system in ArcGIS was the same as my Microstation file and viola it would automatically reference in the exact correct location, without any twisting, warping or moving of the sometimes hundreds of aerial tiles I have to download. 

I have attempted this same approach with AutoCAD, but every single attempt I've made, with changing every conceivable setting in ArcGIS and AutoCAD (including similar settings noted above) the aerial always comes out shifted approximately 12 feet off in one direction or the other.. I have verified there is nothing wrong with the topographic survey by going to the project site and checking the northings and eastings of several points within the survey match up perfectly and have checked against multiple different projects. I have also exported the survey to Microstation and referenced in an aerial in the same fashion noted above and it overlaps near perfectly <0.5 feet off. Can anyone run me through a hopefully straight forward and hopefully simple process of simply getting an aerial photograph to appear in the right location, without wasting hundreds of hours of work on every project and proposal we prepare by having to manually move hundreds of image tiles so the road is actually in the correct location? 

Again, thanks for any help!

---

When your survey was created, was it created in a coordinate system using the survey database? If so, was the survey database coordinate system set correctly? You mentioned that you don't set a coordinate system in your template(s). Before you began inserting any kind of survey data, did you set the coordinate system? The reason I ask is that C3D will not transform objects after insert; only on import. This can be confusing because checking the raw coordinates will seem OK, but the drawing is still off. This happened to me back when Autodesk added US Survey Foot as a unit and not just Foot. 

Starting with Map 3D 2013, Autodesk decided that MAPIINSERT would reproject an image on import. This has led to many issues, such as ignoring an embedded geoheader's UNITS. It sounds like you've got an external worldfile, so that particular issue shouldn't affect you. I only mention this because Autodesk has had bugs with inserting georeferenced images for almost a decade and hasn't fixed them. Perhaps this is one of them?

[pbradburyYWE5N]

We didn't create the survey file, it was provided to us from a third party. And when we received the file, I checked the drawing settings and it had the coordinate system noted. I have this morning tried another topo file and another set of downloaded aerials both in Civil 3D and Microstation with the 2011 NAD 83 US Foot, I even tried NAD 83 2011 meters and none of the aerials come in correct and the discrepancy is the same in Microstation and AutoCAD. I think its ultimately an issue with how I am obtaining the aerials.. 

Your post brought up another question for me... Most firms I would think, would be provided the topo file from a survey company, and their design would be built off this topo file.. When we save-as our template file that doesn't contain a coordinate system, would we be required to set a coordinate system (same as topo file), or if we simply XREF in the topo, the northings and eastings we can get from the X & Y coordinates in the file would be correct regardless? 

[Cadguru42]

---

@pbradburyYWE5N wrote:

We didn't create the survey file, it was provided to us from a third party. And when we received the file, I checked the drawing settings and it had the coordinate system noted. I have this morning tried another topo file and another set of downloaded aerials both in Civil 3D and Microstation with the 2011 NAD 83 US Foot, I even tried NAD 83 2011 meters and none of the aerials come in correct and the discrepancy is the same in Microstation and AutoCAD. I think its ultimately an issue with how I am obtaining the aerials.. 

Your post brought up another question for me... Most firms I would think, would be provided the topo file from a survey company, and their design would be built off this topo file.. When we save-as our template file that doesn't contain a coordinate system, would we be required to set a coordinate system (same as topo file), or if we simply XREF in the topo, the northings and eastings we can get from the X & Y coordinates in the file would be correct regardless? 

---

You always need the coordinate system set first thing before doing any actual work in a drawing. If your drawing did not have the coordinate system set before XREFing/Inserting the survey drawing (that you said has a coordinate system), then C3D will try to adjust it on insert. The default unit is International foot in C3D if no coordinate system is set, so when you inserted/XREF'd the survey DWG that has US Foot, C3D is most likely shifting it automatically. Was the survey drawing an actual C3D drawing with a coordinate system set or just a plain DWG?

I don't think it's the aerials if it's happening in both Microstation and C3D. I have a feeling it's your base drawing isn't in the correct place.

[pbradburyYWE5N]

Thanks for your post, now I feel like I'm getting somewhere! I have a number of questions/clarifications however..

So I make sure I'm understanding correctly, if I..

A. Save-as a template drawing that does NOT have a coordinate system

B. XREF in a drawing that does have a coordinate system, C3D will shift the XREF'd drawing? Or will it shift the drawing I saved-as? I asked this question previously on this forum and was told nothing would happen to either file during this process..

C. When you say INSERT, I am not sure what you mean, are you referring to inserting a block or XREF, XREF versus INSERT are two totally different things?

In terms of, if its the file, I checked another topo file for another project and I get the same discrepancy.. 

[Mike.M.Carlson]

A. Save-as a template drawing that does NOT have a coordinate system

• If your template has no coordinate system defined (i.e., No Datum / No Projection), you still need to check your ACAD UNITS.  Do you want this template file to be in US Survey Feet or International Feet?  IF you try to ATTACH an XREF or INSERT a block that are based on meters, inches, US Survey Feet, Or International feet, ACAD will do a conversion if the template drawing has a different UNIT specified compared to the dwg files you want to bring into this template.  If they are consistent, no conversion will occur.

B. XREF in a drawing that does have a coordinate system, C3D will shift the XREF'd drawing? 

• A conversion will occur in this scenario if the XREF dwg UNITS are different then the template UNITS.  Again, there are two places to check your UNITS...the ACAD UNITS, which is most critical, and the Civil 3D Drawing units defined with a coordinates system (if specified). 

 Or will it shift the drawing I saved-as?

• NO...even if the attached XREF has a coordinate system defined, it will not update your template with this coordinate system after attaching if the template is set to No Datum/No Projection.  However, a conversion will occur during attach, if the XREF ACAD UNITS are different the the template.

I asked this question previously on this forum and was told nothing would happen to either file during this process.

Nothing would happen with respect to a coordinate system update to the template...a conversion will occur upon ATTACH, if the XREF ACAD UNITS are different then you template.

C. When you say INSERT, I am not sure what you mean, are you referring to inserting a block or XREF, XREF versus INSERT are two totally different things?

• INSERT is a term typically used for Blocks, but behaves the same as ATTACH with regard to ACAD UNITS.

[pbradburyYWE5N]

Thanks!

So I'm clear the ACAD units are found under the Insertion Scale, if you simply type in UNITS correct? And the Civil 3D Drawing units are found under the drawing settings for the file I'm in correct? And beyond these two locations, there is not a single other setting, over-ride or other function that can specify or change the coordinates or units of a AutoCAD file correct? 

So, my scenario now is, I have an XREF that when I type UNITS, I get an Insertion Scale of US Survey Feet, these are the ACAD units? 

When I save-as my template and simply type in UNITS and check the Insertion Scale and set it to US Survey Feet, no project or altering should occur if I simply XREF in the other drawing correct? And with the save-as file I don't have to set a coordinate system, or do anything whatsoever with the Civil 3D Drawings units correct? 

And when I XREF, I don't change anything with the Scale, Insertion Point, Rotation or Block Unit and etc.. in the Attach External Reference dialogue correct? 

I have done exactly this with two files I am currently working with and when I check what I assume is a northing and easting readout just below the command line, the same point in the XREF and in the save-as file with the XREF, the readouts are different in both situations...

[Mike.M.Carlson]

So I'm clear the ACAD units are found under the Insertion Scale, if you simply type in UNITS correct?

• Yes that is correct

And the Civil 3D Drawing units are found under the drawing settings for the file I'm in correct?  And beyond these two locations, there is not a single other setting, over-ride or other function that can specify or change the coordinates or units of a AutoCAD file correct? 

• Yes that is correct.  When I refer to Civil 3D drawing units, I mean the UNITs as related to the Coordinate System you choose.  This setting comes into play when using the Survey Data Base or Map 3D connections, etc.
• One caveat though:  If you choose No Datum / No Projection and if you will be working with attaching Metric files, you may also have to toggle the Imperial to Metric Conversion drop down on the Units and Zone Tab.  This is to ensure the subtle difference between US Survey Feet and International Feet are accounted for in a Metric to Imperial conversion.
•   
• The key is to make sure that you specify the ACAD UNITs first.  And if you decide to apply a Coordinate System, make sure the Check Box "Set AutoCAD Variables to Match" is checked...this will sync you ACAD UNITs with the Coordinate System Units you choose. 
•  

So, my scenario now is, I have an XREF that when I type UNITS, I get an Insertion Scale of US Survey Feet, these are the ACAD units?  

• Yes that is correct

When I save-as my template and simply type in UNITS and check the Insertion Scale and set it to US Survey Feet, no project or altering should occur if I simply XREF in the other drawing correct?

• Yes that is correct.  You can confirm if a conversion is occurring upon XREF ATTACH in the XREF attach dialog in lower right hand corner under Scale Factor.  In pick below:  I have a International Foot DWG open and I'm attaching a XREF dwg that has US Survey Foot ACAD UNITs specified (Both have no Datum / No Projection).  Note the Scale factor is greater than 1 when converting US Survey Foot to International Feet.
• 

And with the save-as file I don't have to set a coordinate system, or do anything whatsoever with the Civil 3D Drawings units correct? 

• Yes that is correct, you can specify a Coordinate System or leave it No Datum / No Projection in your template.  However, if you know the UNITs your survey will be in and how you want your drawings set up, then Specify the ACAD UNITs and check the box "Set AutoCAD Variables to Match".  SO if you decide to switch to a published State Plane Systems the ACAD units and Civil 3D Units will sync.  

And when I XREF, I don't change anything with the Scale, Insertion Point, Rotation or Block Unit and etc.. in the Attach External Reference dialogue correct? 

• You may need to increase the significant figures in your template prior to inserting or Attaching XREFs so you can see if the US Survey foot to International Foot conversion is occurring (see image above).  Setting the CAD UNITs to 6 significant figures should show you if a conversion scale factor is being applied upon insertion or attach.

[pbradburyYWE5N]

Thank you very much, very helpful! One last quick question, how does all of this relate to the Base Point 0,0,0, which I think may be similar to the global origin in Microstation, which I have never messed with or changed? I would expect that, possibly the Base Point specifies from which location w/in the file a northing and easting is measured, which I would think if I changed this base point, then I would change the northing and eastings I would get from the same point before hand.. May be getting into the weeds with this, but this was how I fixed the file I'm currently working in, the base point was not set at 0,0,0.. I never had to mess with the global origin in Microstation and I'm unclear why anyone would want to change this in AutoCAD..?

[Mike.M.Carlson]

how does all of this relate to the Base Point 0,0,0, which I think may be similar to the global origin in Microstation, which I have never messed with or changed? I would expect that, possibly the Base Point specifies from which location w/in the file a northing and easting is measured, which I would think if I changed this base point, then I would change the northing and eastings I would get from the same point before hand.. May be getting into the weeds with this, but this was how I fixed the file I'm currently working in, the base point was not set at 0,0,0.. I never had to mess with the global origin in Microstation and I'm unclear why anyone would want to change this in AutoCAD..?

I would not not change this in AutoCAD...I'm actually not even sure you can.  I believe you're referring to the concept of False Northings and False Eastings...These values are set for a given published coordinate system.  @Pointdump please chime in here, as you're the PLS on this thread.   

Base point in AutoCAD or Civil 3D, is the Origin of the Cartesian System in the drawing model.  It can be thought of as Unitless.  No matter what you specify coordinate system-wise or UNITs-wise in a dwg, you can always think of 0,0,0 as the Origin in a given dwg.   For example, if you copied some geometry from one drawing with a base point of 0,0,0 and pasted it into another dwg, you could specify 0,0,1 as the insertion point and it would paste into the new drawing in the exact same location with respect to X and Y, but 1 UNIT higher on Z axis. 

I hope this helps.

[Pointdump]

P,

"...the base point was not set at 0,0,0...I'm unclear why anyone would want to change this in AutoCAD..."

I sure wouldn't want to change the base point of a drawing and can't imagine any reason to do that. Blocks sometimes, but drawings no.

Dave

[chriscowgill7373]

I must have missed something above, in the drawing that was provided, the base point appears to be 0,0,0.  I do agree, though  it is something that should rarely be modified.  even for a block, I tend to move the information as opposed to change the base point, otherwise it just gets confusing.  Had a guy the other day that couldnt figure out why coordinates were different between his base file and the file when it was xreff'ed, come to find out, his base wasnt 0,0,0.  it is not something that presents itself easily and you have to know to look for it, which to me means it is something that one really should never mess with, just like freezing in a viewport (unless you have multiple viewports in a file that you want to see different items)