[Dcmlib] PixelPaddingValue

[Aris Basic]

Message is

Must downscale rescale data from 16bits to 12

i ahvnt really looked if any rescaling actualy gets done but it seams 
like the pixel values get divided by 16 (my assumption thats why i have 
to multiply by 256 instead only by 16 in order to scale the pixel values 
to 16bit)

ill try to get you images tonight (my time) (or put them somwhere where 
you cna download them in order not to transfer that much data
on email servers )

Aris

Jean-Pierre Roux wrote:

> Aris Basic wrote:
>
> [...]
>
>> btw when newer gdcm give out message that pixeldata has to bin 
>> downshift what actualy happens ?
>
>
> --> could you tell us what's the exact message ?
>
> We separate the 'header' analysis and the actual pixel reading.
> We assumed that, in some cases, the user wants to choose the images he 
> wants to work with, just 'looking' at the header.
> Therefore, we don't load 'long'values (>4096 bytes), to save time and 
> RAM space, and we display (at Print time) a message like :
>
> [Pixel Data] [gdcm::NotLoaded Ad.:394226 x(603f2) Lgt:1047272 x(ffae8)]
>
> (But I think you're talking about something else)
>
>> (was looking at gdcm code but i all i've seen is the printing of the 
>> Warn message).
>
>
>
>> (i noticed that with JPEG Lossless compresed 12bit images (packed in 
>> 16) i personaly have to multiply each pixel value with 256 (16*16)
>> and for LEI images its enoigh to multiply it by 16 (in order to 
>> adjust its 12bit range value in 16bit space)
>> (at least that gives me results (on display) comparable to other 
>> dicom viewer [efilm,...] ))
>
>
> It sounds like a bug in gdcm ...
> Could you send me an image with JPEG Lossless compresed 12bit images 
> (packed in 16), to fix the bug.
> Thx
> Jean-Pierre.
>
>>
>> Aris
>>