There is no such thing as printing a raw file directly to CMYK. There exists no demosaicing alorithm that works directly from raw to CMYK.
where/how do you derive these things? do you just make it up? or are you just coming to those realizations?
one has never stated that in any of my commentary. one is fully aware that the RAW is interpolated by the printer driver for the printer.
none of this changes the fact that you can manipulate 16 bit RAW files in a 16 bit environment and send those 16 bit RAW results to the printer to be printed.
i have been doing this for 14 years. i invested in a Canon ipf5000 17" wide format printer in 2006 in order to teach one's self large format printing. i currently print on a Canon Pro4000 44" large format printer. none of this is new. welcome to the real world and the state of the art of large format printing.
Canon’s Digital Photo Professional software interfaces effectively with the printer driver and allows users to select paper profiles, rendering intents and CMYK simulation settings. The printer driver converts the 16 bit RAW RGB print data edited by your application software into CMYK data that your printer can understand and sends the converted data to the printer. The printer driver will do the appropriate conversion from RGB to CMYK to ensure the best results from your digital files.
Direct Printing of Raw Files
In the past, photographers had to convert their raw images into JPEG or TIFF format before they could be printed. However, the growing popularity of the raw file format has encouraged several printer manufacturers to include facilities for printing raw files in some models in their printer range.
software developers have also released applications for printing raw files without the need for an editing interface. There are also Photoshop plug-ins for several high-end printers.
in the real world one should keep it simple. end of line.....
I think you may be using some terms "unconventionally" Once you can see an image, it is no longer "raw", it has been converted to an RGB file in some colour space and displayed on your (hopefully, calibrated) monitor. It can then be "interpreted" by a printer driver (after editing, correction, profiling etc) or it can be saved as an RGB file or a jpeg or TIFF etc. Opening a Canon raw file in DPP requires conversion to a (no longer "raw") image that can be displayed/edited and eventually sent to a printer.
InnomnateViem wrote:
where/how do you derive these things? do you just make it up? or are you just coming to those realizations?
one has never stated that in any of my commentary. one is fully aware that the RAW is interpolated by the printer driver for the printer.
none of this changes the fact that you can manipulate 16 bit RAW files in a 16 bit environment and send those 16 bit RAW results to the printer to be printed.
i have been doing this for 14 years. i invested in a Canon ipf5000 17" wide format printer in 2006 in order to teach one's self large format printing. i currently print on a Canon Pro4000 44" large format printer. none of this is new. welcome to the real world and the state of the art of large format printing.
Canon’s Digital Photo Professional software interfaces effectively with the printer driver and allows users to select paper profiles, rendering intents and CMYK simulation settings. The printer driver converts the 16 bit RAW RGB print data edited by your application software into CMYK data that your printer can understand and sends the converted data to the printer. The printer driver will do the appropriate conversion from RGB to CMYK to ensure the best results from your digital files.
Direct Printing of Raw Files
In the past, photographers had to convert their raw images into JPEG or TIFF format before they could be printed. However, the growing popularity of the raw file format has encouraged several printer manufacturers to include facilities for printing raw files in some models in their printer range.
software developers have also released applications for printing raw files without the need for an editing interface. There are also Photoshop plug-ins for several high-end printers.
in the real world one should keep it simple. end of line.....
I can also "print directly from raw" with CaptureOne and an Epson printer without any special workflow or software. This is not a Canon feature, and it does not change the way conversions are done. It is a very simple integration feature that just saves you the step of using disk space for the TIFF or jpg file.
If you still believe that your files are not going through a doube conversion, first from RAW to RBG color space, and then from RGB to CMYK / printer color space, then you are not paying attention to the details you can read in the sources referenced. You don't even seem to be aware of or understand the difference between demosaic and interpolation. Hey, you don't even seem to understand the difference between a raw file and an RGB file, as you speak about RAW RGB files being converted directly to CMYK.
i see. arrogant condescending strawman arguments, subjective contradictive diatribe, and veiled insults and on top of that you're arguing for arguments sake with a superiority complex.
hmmm, gentlemen surely you 2 can do better than that? you 2 are to damned preoccupied with attempting to prove me wrong. but you can't now can can you? you are not qualified to do so.
and the fact still remains 14 bit RAW capture, 16 bit RAW editing, 16 bit RAW printing. what will the R5's bit depth be?
There will be an answer to your question, at the time the R5 is actually announced.
Since you asked the question here, you get more information than you asked for, and really useful information. That information is that is is counterproductive to consider a camera with lower resolution and higher bit depth even when printing with a 16 bits printer driver. This is especially true when you have a wonderful large 44" printer which actually can show all the extra resolution.
The most important reason is, as said before by not just 2 arrogant gentlemen, that the difference you see between printing 8 bits and 16 bilts files, is thanks to the very few first bits more than 8, like up to 10 bits. This is because ink dots on paper is not able to provide more tonal resolution than you get from 10 bits. It is of course a bit more complicated, as bit depth in the RAW file does not translate directly to effective bit depth in the RBG image sent to the printer driver. Many people even say you can't see any difference above 8 bits, but I will give you those two bits as you can actually see a difference yourself.
If you really want to choose a camera in order to maximize the tonality in your images, you should be more concerned about the full well capacity / maximum SNR / tonal range in the camera sensor, normalized for the sensor resolution. That metric is usually not correlated to the raw bit depth at all. The difference between two sensors both with 14 bits raw files can be significant.
You may of course do whatever you want, but we are here to help.
alundeb wrote:
Many people even say you can't see any difference above 8 bits, but I will give you those two bits as you can actually see a difference yourself.
I think the reason he saw a difference between 8 bit output and 16 bit output is because he used an 8 bit workflow to prepare the 8 bit print as well. Of course various adjustments in that workflow could introduce visible artifacts like gradient banding. Not only that, but this also means the adjustments may have differed for the two images being printed.
A proper comparison would be to prepare an image using a 16 bit workflow, then print the result twice — once by sending 16 bit data to the printer, and once sending only 8 bit data to the printer. This is easy to do with Canon’s printer driver from within Photoshop, as the print dialog contains a check box which says “send 16 bit data to printer.“ If it isn’t checked, it sends only 8 bit data.
Looking back on this discussion, I am somewhat confused by a few things which, taken individually, did not raise any red flags. But taken together, they are worth addressing. Those things are: 1) The OP is concerned about the native bit depth captured by the sensor; 2) This concern arises because he has seen better output with a 16 bit workflow than with an 8 bit workflow, and concludes that more bits are better; 3) He’s primarily concerned with DSO astrophotography; and 4) He seems to be asserting that he does no processing of images in Photoshop or any image editing software other than DPP and Print Studio Pro, maintaining a “16 bit Canon RAW workflow.”
If one is really concerned about DSO image quality, they would see far more benefit to image quality by stacking a bunch of subs than by worrying oneself with whether the sensor captures 10, 12, or 14 bits/channel and then utilizing a workflow which precludes stacking. (Unless DPP or PSP allows stacking, which is news to me if true.) Stacking requires that images be converted to a format other than RAW, even if the stacking software accepts RAW files as input.
In this case, the most significant impediment to high quality printed images is not whether the sensor captures 10, 12, or 14 bit/channel data, but the methods being used during capture and the use or avoidance of various post-processing tools.
under no circumstances do i work in 8 bits. however, when testing the Pro-4000 i did send both 8 and 16 bit data for a given image to the printer for comparison.
He seems to be asserting that he does no processing of images in Photoshop or any image editing software other than DPP and Print Studio Pro, maintaining a “16 bit Canon RAW workflow.”
this is true in reference to standard single shot imaging. but atrophotography demands a radically different workflow.
individual subs can be edited before stacking within specific parameters ie. the data must remain linear for further PP to avoid artifacts.
a brief astro work flow:
1. stack RAW subs in DSS without WB correction in DSS. the result will be a linear 16 bit stacked FITS or TIFF image with embedded stretched data.
2. open DSS TIFF file in StarTools for PP. develop the composite. the result will be a stacked stretched TIFF image. save file.
3. fine tune developed composite TIFF image. open image in DPP or DxO PhotoLab 3 dependent upon the specifics of fine tuning.
because no 3rd party software to my knowledge, can read Canon's DPP edit data. one is forced to excluded DPP in the initial or pre-stacked RAW workflow.
If one is really concerned about DSO image quality, they would see far more benefit to image quality by stacking a bunch of subs... Stacking requires that images be converted to a format other than RAW, even if the stacking software accepts RAW files as input.
true to a certain degree. longer subs > 5 min retain more colour fidelity, until a point of diminishing returns. agreed that subs are better in most instances, not including single long exposure dark site imaging.
Once you can see an image, it is no longer "raw", it has been converted to an RGB file in some colour space and displayed on your (hopefully, calibrated) monitor. It can then be "interpreted" by a printer driver (after editing, correction, profiling etc) or it can be saved as an RGB file or a jpeg or TIFF etc. Opening a Canon raw file in DPP requires conversion to a (no longer "raw") image that can be displayed/edited and eventually sent to a printer.