High resolution problem less than 130mm zoom

[[Ti...]]

Once we set our scan at 120mm or less zoom, we get very poor surface quality on the volume and polygonized mesh if we use full resolution. If we use half resolution the volume and scan looks absolutley amazing!!!

 

We are better off scanning 2 parts at 60 zoom at half resolution than 4 parts full resolution at 120 zoom. That works for small parts, but for larger parts that need to be moved out to 120mm zoom we get poor surfaces. The same part pushed back to 300mm zoom will look amazing, but we lose resolution at 300mm obviously.

 

We have many .05mm tolerance surface profiles and scan quality is critical for us.

 

Once we go to 60mm zoom we cannot consider using full resolution if our part has surface profile tolerances, we must use half resolution...Could someone that works in the xray department comment on this issue? We've been using the Metrotom for 4 years and tried everything you can imagine...we've tried no filter, .5Cu, 1mm Cu, 1000-3000 images, 200-1200ms, hardening on/off, etc, volume filtering off/on, etc. This has to be a known issue.

 

Thanks

 

Edited April 19

[[DW...]]

Please sign in to view this username.

 The attachment is from the Tech Guide in the Zeiss Quality Suite. I never use full resolution, because the software is assigning a grey value to every single pixel. As I understand it, this is why you get the rippled surface effect. My go to for acqusition is to use the suggested full resolution images (for example 3000) and then change the quality to half resolution. This way I have plenty of images available for a high quality volume reconstruction, but it then uses 4 pixels (2 in each direction) for that same gray value as it does in high resolution. This strategy has been much more representative of the actual surface. By taking double the amount of recommended images for half resolution, I also find I don't lose some of my smallest features. Play around with it and see if you can get similar results.

[[Ti...]]

Thanks, I'm trying that today as we speak on a stainless part. We have a couple high res scans that are awesome, but they are at 130mm and the part geometry is strange. A more simple part at 130mm high res has the bad surface quality...that always was a head scratcher. Ive tried many different part angles in the fixture, and just gave up and later learned that high res is just problematic closer to the beam. 

 

Have you ever tried to figure out the difference in quality from something like 2500 images @ 600ms vs 3000 images @500ms.  Both add up to a 25 min scan, but I've never come to a conclusion on whether I'd go more images or more exposue time if I had to choose one over the other. 

[[DW...]]

Please sign in to view this quote.

Please sign in to view this username.

 I have not, but in your example I cannot imagine it makes too much of a difference. I am usually never worried about cycle times, so I usually opt for the longer scan.

[[Ti...]]

Please sign in to view this username.

  I tend to agree with you somewhat on the difference I explained.  But 500 images is significant, specially if a scan you were trying to improve was at 1500 images at 500ms...you could go to 2000 images at 400ms and be at nearly the same time.. I'm testing that right now.  I'm wondering if the exposure time is more important with the denser materials and the images is more important for accuracy on the less dense materials.  I'm looking for really getting lean and mean on our scans, I'd like to know if I should lean more on exposure time or images if I'm trying to get the best of the best scan.