Cocoa mail archive
Success Re: Why does my PDF NSImage print to paper at low res screen resolution? (was Re: how to get a sub piece of an NSImage or make a new NSImage that is a subset of an existing one)
FROM : Chad Leigh
DATE : Sat Aug 12 19:33:27 2006
On Aug 12, 2006, at 2:59 AM, Chad Leigh wrote:
>
> On Aug 12, 2006, at 1:47 AM, Chad Leigh wrote:
>
>>
>> On Aug 12, 2006, at 12:01 AM, Chad Leigh wrote:
>>
>>>
>>> On Aug 11, 2006, at 11:43 PM, Jonathan Grynspan wrote:
>>>
>>>> You can find more information on Cocoa's graphics model at
>>>> http://developer.apple.com/documentation/Cocoa/Conceptual/
>>>> CocoaDrawingGuide/index.html . Of particular interest to you may
>>>> be the sections on Graphics Contexts and Images.
>>>
>>> I am reading that now, thanks.
>>>
>>> Unfortunately my new image is a cached screen representation and
>>> not the hi-res PDF that it was before. And it appears from the
>>> CocoaDrawingGuide docs under "Creating Graphics Contexts" that
>>> you cannot create a new PDF image using the code below. It
>>> says :-( "Important: You cannot create a viable graphics context
>>> for PDF or PostScript canvases using the
>>> graphicsContextWithAttributes: method. You must go through the
>>> Cocoa Printing system instead." That however, is what I need to
>>> do, I think.
>>
>> It looks like something like this works:
>>
>> NSImageView *imageView = [[[NSImageView alloc]
>> initWithFrame:labelPDFRect] autorelease];
>> [imageView setImage:newLabel];
>> [imageView beginDocument];
>> NSMutableData *imageData = [NSMutableData data];
>> NSPrintOperation *pdfOp = [NSPrintOperation
>> PDFOperationWithView:imageView insideRect:imageRect
>> toData:imageData];
>> [pdfOp runOperation];
>> label = [[NSImage alloc] initWithData:imageData];
>> // [pdfOp cleanUpOperation];
>> [imageView endDocument];
>>
>>
>> I get my image and it is still a PDF based image by looking at the
>> single NSImageRep (NSPDFImageRep) that it contains, though I have
>> not yet gotten my printing code to spit out the image as I seem to
>> have screwed something up and all I get now is a gray box the size
>> of the page (without margins) so the paper comes out with a gray
>> box with normal white margin. This happens even when I use my
>> PNG image code so I screwed something up somewhere.
>
> Ok, I fixed my screwed up print code -- I had a blank subview that
> had been added in one of my dead-end attempts at fixing this.
>
> It appears that I am getting a PDF image from the above code.
> However, it is either a screen image quality representation or when
> I draw it to the printer view in my NSPrintOperation, it somehow
> gets converted to a screen resolution low res.
>
> I capture the PDF to a file when I get it from the wire and I can
> print that and it is a "high resolution" image just like you would
> expect the PDF to be.
>
> I have an alternate version of my code that uses PDFPage/PDFView
> etc with AffineTransforms to print to my page from just a portion
> of the PDF page. So I know the original PDF is not a low res
> bitmap to start
>
> Here is how I create the original NSIMage
>
> [finalImage appendBytes:(const void *)finalImageBytes length:k];
> // [finalImage writeToFile:@"express_label.pdf" atomically:YES];
> newLabel = [[NSImage alloc] initWithData:finalImage];
>
> (gdb) po newLabel
> NSImage 0x1607ba30 Size={612, 792} Reps=(
> NSPDFImageRep 0x1609ac40 Size={612, 792}
> ColorSpace=NSCalibratedRGBColorSpace BPS=0 Pixels=612x792 Alpha=NO
> )
> (gdb)
>
> Question on the above -- is the Pixels here a real pixels size or
> is it like the page -- a coordinate system that does not really
> relate to actual pixels? The way the image is drawn it appears to
> be a real pixel count -- how can I get the NSImage (NSPDFImageRep)
> to not degrade when I create it?
>
> // (I now calculate some rects to grab the top half of the image
> and make a new image -- this part is snipped )
>
> // create new image
>
> NSImageView *imageView = [[[NSImageView alloc]
> initWithFrame:labelPDFRect] autorelease];
> [imageView setImage:newLabel];
> [imageView beginDocument];
> NSMutableData *imageData = [NSMutableData data];
> NSPrintOperation *pdfOp = [NSPrintOperation
> PDFOperationWithView:imageView insideRect:imageRect toData:imageData];
> [pdfOp runOperation];
> label = [[NSImage alloc] initWithData:imageData];
> // [pdfOp cleanUpOperation];
> [imageView endDocument];
>
> (gdb) po label
> NSImage 0x1685f80 Size={612, 416} Reps=(
> NSPDFImageRep 0x16063a30 Size={612, 416}
> ColorSpace=NSCalibratedRGBColorSpace BPS=0 Pixels=612x416 Alpha=NO
> )
> (gdb)
>
>
> return [label autorelease];
>
> // gets saved away
>
> // later in print code
>
> // calculate a new target rect based on the paper size from the
> NSPrintInfo -- not shown
>
> // print to the print view -- show in gdb the label before it is drawn
>
> (gdb) po label
> NSImage 0x1685f80 Size={612, 416} Reps=(
> NSPDFImageRep 0x16063a30 Size={612, 416}
> ColorSpace=NSCalibratedRGBColorSpace BPS=0 Pixels=612x416 Alpha=NO
> )
> (gdb)
>
> [label drawInRect:labelRect fromRect:imageRect
> operation:NSCompositeCopy fraction:1.0];
>
>
> // more drawing code for text etc
>
> When this gets printed on paper the PDF is a low-res "screen" type
> image and not the crisp original PDF image
Ok, even though my NSImage only has 1 representation, the
NSPDFImageRep, it draws a low-res screen copy. But Andrew Stone to
the rescue!
I found on this page <http://stone.com/The_Cocoa_Files/
Cocoamotion.html> the answer (look for PDF Optimization). He says
that you must take the actual NSPDFImageRep out and cause it to draw
itself. His code looks like
if (needToUseRealPDFData) {
// grab the image's best representation, the actual
NSPDFImageRep:
NSPDFImageRep *rep = [_image bestRepresentationForDevice:nil];
// ask it to draw within our bounds:
[rep drawInRect:NSMakeRect
(0.0,0.0,_bounds.size.width,_bounds.size.height)];
} else
[image drawInRect:NSMakeRect
(0.0,0.0,_bounds.size.width,_bounds.size.height) fromRect:NSZeroRect
operation:NSCompositeSourceOver fraction:1.0 - _dissolve];
which for testing purposes in my code looks like (I'll clean it up
and set up a pref so I can use PNG or PDF depending on what the user
wants and what comes over the wire, etc)
if (1) {
// grab the image's best representation, the actual
NSPDFImageRep:
NSPDFImageRep *rep = [label bestRepresentationForDevice:nil];
// ask it to draw within our bounds:
[rep drawInRect:labelRect];
}
else
[label drawInRect:labelRect fromRect:imageRect
operation:NSCompositeCopy fraction:1.0];
Thanks Andy!
I've also modified my code for extracting the piece of PDF I need to
use the NSView helper routine - dataWithPDFInsideRect:aRect; since it
seemed easier and I let it set up the print op etc.
Chad
>
> What have I done wrong?
>
> Thanks
> Chad
>
>
>>
>> Chad
>>
>>
>>>
>>> Thanks
>>> Chad
>>>
>>>>
>>>> -Jonathan Grynspan
>>>>
>>>> On 12-Aug-06, at 1:34 AM, Chad Leigh wrote:
>>>>
>>>>>
>>>>> On Aug 11, 2006, at 11:28 PM, Jonathan Grynspan wrote:
>>>>>
>>>>>> Try the following:
>>>>>>
>>>>>> NSImage *partOfImage(NSImage *input, NSRect targetRect) {
>>>>>> if (input) {
>>>>>> NSImage *output = [[NSImage alloc] initWithSize:
>>>>>> targetRect.size];
>>>>>> [output lockFocus];
>>>>>> [input
>>>>>> drawAtPoint: NSZeroPoint
>>>>>> fromRect: targetRect
>>>>>> operation: NSCompositeCopy
>>>>>> fraction: 1.0f];
>>>>>> [output unlockFocus];
>>>>>> return [output autorelease];
>>>>>> } else {
>>>>>> return nil;
>>>>>> }
>>>>>> }
>>>>>
>>>>> Thanks, I will. I've been reading the lockFocus and the
>>>>> drawAtPoint stuff since I posted the first request, trying to
>>>>> put my brain around it. This helps immensely!
>>>>>
>>>>> Thanks
>>>>> Chad
>>>>>
>>>>>>
>>>>>> -Jonathan Grynspan
>>>>>>
>>>>>> On 12-Aug-06, at 1:17 AM, Chad Leigh wrote:
>>>>>>
>>>>>>> I have a Cocoa app and I get a PDF image over the internet
>>>>>>> that I store in an NSImage using
>>>>>>>
>>>>>>> label = [[NSImage alloc] initWithData:finalImage];
>>>>>>>
>>>>>>>
>>>>>>> The image previously was a PNG image and was exactly the
>>>>>>> image needed. Now that they pass a PDF image, they pass an
>>>>>>> image that is an image of a complete piece of paper with the
>>>>>>> image I am interested in placed on the upper half of this
>>>>>>> virtual paper (the rest is just blank). The image is always
>>>>>>> in the same place and the same size so I can determine a Rect
>>>>>>> that will always work to select it. I want to extract this
>>>>>>> image out of the PDF and create a new NSImage (or overwrite
>>>>>>> the old one, I don't care) that I can use later on to print
>>>>>>> in my own existing print routine and to display thumbnails of
>>>>>>> it.
>>>>>>>
>>>>>>> I do not see how to go about it. Most of the draw/composite
>>>>>>> routines I see in NSImage rely on an existing userspace to
>>>>>>> draw into. I want to "draw" into a new NSImage. My brain is
>>>>>>> having a hard time adjusting to think about how to do this.
>>>>>>>
>>>>>>> Ideally I want something like
>>>>>>>
>>>>>>> NSImage *newImage = [[NSImage alloc] initFromImage:oldImage
>>>>>>> usingRect:targetRect];
>>>>>>>
>>>>>>> I expect what I want to do is very easy but if someone could
>>>>>>> point me on the way I should be exploring I would appreciate it.
>>>>>>>
>>>>>>>
>>>>>>> Thanks
>>>>>>> Chad
>>>>>>>
DATE : Sat Aug 12 19:33:27 2006
On Aug 12, 2006, at 2:59 AM, Chad Leigh wrote:
>
> On Aug 12, 2006, at 1:47 AM, Chad Leigh wrote:
>
>>
>> On Aug 12, 2006, at 12:01 AM, Chad Leigh wrote:
>>
>>>
>>> On Aug 11, 2006, at 11:43 PM, Jonathan Grynspan wrote:
>>>
>>>> You can find more information on Cocoa's graphics model at
>>>> http://developer.apple.com/documentation/Cocoa/Conceptual/
>>>> CocoaDrawingGuide/index.html . Of particular interest to you may
>>>> be the sections on Graphics Contexts and Images.
>>>
>>> I am reading that now, thanks.
>>>
>>> Unfortunately my new image is a cached screen representation and
>>> not the hi-res PDF that it was before. And it appears from the
>>> CocoaDrawingGuide docs under "Creating Graphics Contexts" that
>>> you cannot create a new PDF image using the code below. It
>>> says :-( "Important: You cannot create a viable graphics context
>>> for PDF or PostScript canvases using the
>>> graphicsContextWithAttributes: method. You must go through the
>>> Cocoa Printing system instead." That however, is what I need to
>>> do, I think.
>>
>> It looks like something like this works:
>>
>> NSImageView *imageView = [[[NSImageView alloc]
>> initWithFrame:labelPDFRect] autorelease];
>> [imageView setImage:newLabel];
>> [imageView beginDocument];
>> NSMutableData *imageData = [NSMutableData data];
>> NSPrintOperation *pdfOp = [NSPrintOperation
>> PDFOperationWithView:imageView insideRect:imageRect
>> toData:imageData];
>> [pdfOp runOperation];
>> label = [[NSImage alloc] initWithData:imageData];
>> // [pdfOp cleanUpOperation];
>> [imageView endDocument];
>>
>>
>> I get my image and it is still a PDF based image by looking at the
>> single NSImageRep (NSPDFImageRep) that it contains, though I have
>> not yet gotten my printing code to spit out the image as I seem to
>> have screwed something up and all I get now is a gray box the size
>> of the page (without margins) so the paper comes out with a gray
>> box with normal white margin. This happens even when I use my
>> PNG image code so I screwed something up somewhere.
>
> Ok, I fixed my screwed up print code -- I had a blank subview that
> had been added in one of my dead-end attempts at fixing this.
>
> It appears that I am getting a PDF image from the above code.
> However, it is either a screen image quality representation or when
> I draw it to the printer view in my NSPrintOperation, it somehow
> gets converted to a screen resolution low res.
>
> I capture the PDF to a file when I get it from the wire and I can
> print that and it is a "high resolution" image just like you would
> expect the PDF to be.
>
> I have an alternate version of my code that uses PDFPage/PDFView
> etc with AffineTransforms to print to my page from just a portion
> of the PDF page. So I know the original PDF is not a low res
> bitmap to start
>
> Here is how I create the original NSIMage
>
> [finalImage appendBytes:(const void *)finalImageBytes length:k];
> // [finalImage writeToFile:@"express_label.pdf" atomically:YES];
> newLabel = [[NSImage alloc] initWithData:finalImage];
>
> (gdb) po newLabel
> NSImage 0x1607ba30 Size={612, 792} Reps=(
> NSPDFImageRep 0x1609ac40 Size={612, 792}
> ColorSpace=NSCalibratedRGBColorSpace BPS=0 Pixels=612x792 Alpha=NO
> )
> (gdb)
>
> Question on the above -- is the Pixels here a real pixels size or
> is it like the page -- a coordinate system that does not really
> relate to actual pixels? The way the image is drawn it appears to
> be a real pixel count -- how can I get the NSImage (NSPDFImageRep)
> to not degrade when I create it?
>
> // (I now calculate some rects to grab the top half of the image
> and make a new image -- this part is snipped )
>
> // create new image
>
> NSImageView *imageView = [[[NSImageView alloc]
> initWithFrame:labelPDFRect] autorelease];
> [imageView setImage:newLabel];
> [imageView beginDocument];
> NSMutableData *imageData = [NSMutableData data];
> NSPrintOperation *pdfOp = [NSPrintOperation
> PDFOperationWithView:imageView insideRect:imageRect toData:imageData];
> [pdfOp runOperation];
> label = [[NSImage alloc] initWithData:imageData];
> // [pdfOp cleanUpOperation];
> [imageView endDocument];
>
> (gdb) po label
> NSImage 0x1685f80 Size={612, 416} Reps=(
> NSPDFImageRep 0x16063a30 Size={612, 416}
> ColorSpace=NSCalibratedRGBColorSpace BPS=0 Pixels=612x416 Alpha=NO
> )
> (gdb)
>
>
> return [label autorelease];
>
> // gets saved away
>
> // later in print code
>
> // calculate a new target rect based on the paper size from the
> NSPrintInfo -- not shown
>
> // print to the print view -- show in gdb the label before it is drawn
>
> (gdb) po label
> NSImage 0x1685f80 Size={612, 416} Reps=(
> NSPDFImageRep 0x16063a30 Size={612, 416}
> ColorSpace=NSCalibratedRGBColorSpace BPS=0 Pixels=612x416 Alpha=NO
> )
> (gdb)
>
> [label drawInRect:labelRect fromRect:imageRect
> operation:NSCompositeCopy fraction:1.0];
>
>
> // more drawing code for text etc
>
> When this gets printed on paper the PDF is a low-res "screen" type
> image and not the crisp original PDF image
Ok, even though my NSImage only has 1 representation, the
NSPDFImageRep, it draws a low-res screen copy. But Andrew Stone to
the rescue!
I found on this page <http://stone.com/The_Cocoa_Files/
Cocoamotion.html> the answer (look for PDF Optimization). He says
that you must take the actual NSPDFImageRep out and cause it to draw
itself. His code looks like
if (needToUseRealPDFData) {
// grab the image's best representation, the actual
NSPDFImageRep:
NSPDFImageRep *rep = [_image bestRepresentationForDevice:nil];
// ask it to draw within our bounds:
[rep drawInRect:NSMakeRect
(0.0,0.0,_bounds.size.width,_bounds.size.height)];
} else
[image drawInRect:NSMakeRect
(0.0,0.0,_bounds.size.width,_bounds.size.height) fromRect:NSZeroRect
operation:NSCompositeSourceOver fraction:1.0 - _dissolve];
which for testing purposes in my code looks like (I'll clean it up
and set up a pref so I can use PNG or PDF depending on what the user
wants and what comes over the wire, etc)
if (1) {
// grab the image's best representation, the actual
NSPDFImageRep:
NSPDFImageRep *rep = [label bestRepresentationForDevice:nil];
// ask it to draw within our bounds:
[rep drawInRect:labelRect];
}
else
[label drawInRect:labelRect fromRect:imageRect
operation:NSCompositeCopy fraction:1.0];
Thanks Andy!
I've also modified my code for extracting the piece of PDF I need to
use the NSView helper routine - dataWithPDFInsideRect:aRect; since it
seemed easier and I let it set up the print op etc.
Chad
>
> What have I done wrong?
>
> Thanks
> Chad
>
>
>>
>> Chad
>>
>>
>>>
>>> Thanks
>>> Chad
>>>
>>>>
>>>> -Jonathan Grynspan
>>>>
>>>> On 12-Aug-06, at 1:34 AM, Chad Leigh wrote:
>>>>
>>>>>
>>>>> On Aug 11, 2006, at 11:28 PM, Jonathan Grynspan wrote:
>>>>>
>>>>>> Try the following:
>>>>>>
>>>>>> NSImage *partOfImage(NSImage *input, NSRect targetRect) {
>>>>>> if (input) {
>>>>>> NSImage *output = [[NSImage alloc] initWithSize:
>>>>>> targetRect.size];
>>>>>> [output lockFocus];
>>>>>> [input
>>>>>> drawAtPoint: NSZeroPoint
>>>>>> fromRect: targetRect
>>>>>> operation: NSCompositeCopy
>>>>>> fraction: 1.0f];
>>>>>> [output unlockFocus];
>>>>>> return [output autorelease];
>>>>>> } else {
>>>>>> return nil;
>>>>>> }
>>>>>> }
>>>>>
>>>>> Thanks, I will. I've been reading the lockFocus and the
>>>>> drawAtPoint stuff since I posted the first request, trying to
>>>>> put my brain around it. This helps immensely!
>>>>>
>>>>> Thanks
>>>>> Chad
>>>>>
>>>>>>
>>>>>> -Jonathan Grynspan
>>>>>>
>>>>>> On 12-Aug-06, at 1:17 AM, Chad Leigh wrote:
>>>>>>
>>>>>>> I have a Cocoa app and I get a PDF image over the internet
>>>>>>> that I store in an NSImage using
>>>>>>>
>>>>>>> label = [[NSImage alloc] initWithData:finalImage];
>>>>>>>
>>>>>>>
>>>>>>> The image previously was a PNG image and was exactly the
>>>>>>> image needed. Now that they pass a PDF image, they pass an
>>>>>>> image that is an image of a complete piece of paper with the
>>>>>>> image I am interested in placed on the upper half of this
>>>>>>> virtual paper (the rest is just blank). The image is always
>>>>>>> in the same place and the same size so I can determine a Rect
>>>>>>> that will always work to select it. I want to extract this
>>>>>>> image out of the PDF and create a new NSImage (or overwrite
>>>>>>> the old one, I don't care) that I can use later on to print
>>>>>>> in my own existing print routine and to display thumbnails of
>>>>>>> it.
>>>>>>>
>>>>>>> I do not see how to go about it. Most of the draw/composite
>>>>>>> routines I see in NSImage rely on an existing userspace to
>>>>>>> draw into. I want to "draw" into a new NSImage. My brain is
>>>>>>> having a hard time adjusting to think about how to do this.
>>>>>>>
>>>>>>> Ideally I want something like
>>>>>>>
>>>>>>> NSImage *newImage = [[NSImage alloc] initFromImage:oldImage
>>>>>>> usingRect:targetRect];
>>>>>>>
>>>>>>> I expect what I want to do is very easy but if someone could
>>>>>>> point me on the way I should be exploring I would appreciate it.
>>>>>>>
>>>>>>>
>>>>>>> Thanks
>>>>>>> Chad
>>>>>>>
