Issues with Extended Offset Table

[lokeshreddyvaka]

I am using the pyidcom's encapsulated_extended method to get the encapsulted pixel data & its corresponding extended offset table & extended offset table lenghts. But I think that the offsets in the Extended Offset Table seem to be one byte less than what the lengths of Pixel Data frames suggest. I used dcmdump tool to read the metadata. Attaching the screenshot below,

As per the pixel data, the length of first frame is 1652 bytes, so the byte offset for 2nd frame should be 1660(1652+8 for the tags and length), but in the Extended Offset table(7fe0, 0001), it mentioned that the byte offset is 1659(i.e the start position of 2nd frame). Also interestingly in the Extended Offset table lengths(7fe0, 0002), it is mentioned that the length of first byte is 1651 bytes. I am not sure why there is difference in length for first frame. Any help would be appreciated! Thanks

[scaramallion]

Thanks for letting us know, this was fixed with #1970.

[lokeshreddyvaka]

@scaramallion Thank you for your response. I would like to clarify if, with version "2.3.1," the issue was limited to the extended offset table or if it also affected the pixel data.

[lokeshreddyvaka]

@scaramallion I regenerated the dicoms with upgraded version of pydicom library(2.4.4), however I am encountering the same issues as before --> As per the pixel data, the length of first frame is 1652 bytes, but in the Extended Offset table lengths(7fe0, 0002), it is mentioned that the length of first byte is 1651 bytes.
upgrading the version didn't solve my problem, is there anything that I have to do apart from this?

[scaramallion]

2.4.4 still has the issue, the fix won't be live until the v3.0 release.

And yes, this only affects the extended offset table not the actual pixel data.

[scaramallion]

Until then you can implement your own local version of encapsulate_extended()

[lokeshreddyvaka]

Thanks @scaramallion