I am developing an application to read an image format and convert it to my proprietary format. I want to read/get the layer information of the image, each graphic object associated with the image.
I downloaded CS SDK but unable to identify the component/APIs. Can you please help me in identifying the same.
So, you are looking to automate Photoshop from your application, to
have it do the task? Or you're looking to add on a Photoshop format
plug-in to have it export to your proprietary format (which sounds a
better fit for the SDK).
You'll need the "Advanced" SDK to find out anything about PSD/PSB, but the documentation is rather incomplete. http://telegraphics.com.au/svn/psdparse/trunk is source code for a program to parse it - at least, that part that was documented sufficiently or I could figure out.
I want to read Bartender (CD labelling image) format and create my own CD label image. Photoshop is able to load the bartender file and also allows to get the layers from it.
So, I thought I could use photoshop SDK to load the file and get the graphic information from the file and create my own format.
Maybe you can use the Photoshop SDK, but you don't make it clear
whether you completely understand what the SDK is. It is a set of
tools for adding on to Photoshop, or automating Photoshop.
The "obvious" way to do this is to add an export module to Photoshop,
so people can save anything open in Photoshop into your file format.
What do you mean by "graphic information", though - do you just mean
the raster images of each layer?
Hi all,
I am using PSD file documentation to make the psd parser.. I would like to ask inside the Layer and Mask Information section, there is line written on Channel Image Data i.e. "If the layer's size, and therefore the data, is odd, a pad byte will be inserted at the end of the row"
What is meant by this line exactly and what i have to do in this case..
I am reading it in to the buffer..
I don't have that document, but this is a very common feature of image
file formats. If so...
Image data is stored in rows, so the first piece of data is the first
row (top or bottom), the second piece of data is the second row, and
so on.
What this is saying is that each row will start on an even numbered
byte. So if the rows occupy, say, 224 bytes, they will follow each
other exactly. But if the rows occupy 225 bytes there is an extra
byte. Each row uses 226 bytes, but only the first 225 bytes are used.
Some image formats have padding only to a byte boundary (e.g. PDF
images), some might have other rules (e.g. BMP pads to a multiple of 4
bytes).
> I am using PSD file documentation to make the psd parser..
If you want an already written parser, try my http://telegraphics.com.au/sw/#psdparse - Trying to puzzle out all the ambiguities and omissions of the PSD documentation is an unrewarding few hundred hours, good luck.
Actually i have tried all this by adding a byte in my destination buffer and i have skipped 1 byte in my source buffer but neither of them worked..in getting the image when data is odd(i am getting image when data is even).
fyi http://www.fileformat.info/format/psd/egff.htm
I think you have given me a plug in for a psd file. but i need the logic to solve this problem.
Hi Aandi,
While reading the source buffer i am storing bytes into the integer buffer and after reading all the channels in to separate buffers, i am combining all the channel buffers in to single surface(Directx).
I don't know anything about DirectX, but what are its row padding
requirements?
Also, have you tried looking directly at the data returned for the
layers to see whether it matches your expectation? A small white
bitmap with a 1-pixel wide left border is a good starting point.
Hi Aandi,
I have checked, there is no row padding requirements for the surface in DirectX. and i have tried your method by adding a bitmap of one pixel but the result was same(corrupt image)......I have tried all the possibilities to decode the line("If the layer's size, and therefore the data, is odd, a pad byte will be inserted at the end of the row") but always the result was same..
I need two important information regarding the PSD..
1.Is it possible to extract the RGB information of each layer using..If so how??
2.If i draw a simple Circle in adobe photshop and save it as hello.psd.Later while extracting the hello.psd using the psd parser, how will i know that the image is a circle??
16
Replies
AdChoices