16bit vs 8bit printing on 9900

[Garnick]

Is there really a noticeable/observable difference?  I think I have seen this covered here, but couldn't seem to find the post(s).  A customer has just dumped a load of 16bit files for printing and I really don't want to get bogged down sending 500Mb files to the printer.  I'm not getting paid enough for the hassle.  Any replies will be greatly appreciated.

Gary

[Slobodan Blagojevic]

I believe the consensus here is that there is no practical difference. Perhaps a very trained eye, looking with a very strong loupe, might detect a slight difference.

[Bart_van_der_Wolf]

Is there really a noticeable/observable difference?  I think I have seen this covered here, but couldn't seem to find the post(s).  A customer has just dumped a load of 16bit files for printing and I really don't want to get bogged down sending 500Mb files to the printer.  I'm not getting paid enough for the hassle.  Any replies will be greatly appreciated.

Hi Gary,

The differences will usually be very small, perhaps only detectable in a direct side-by-side comparison with a loupe.

The 16-bits will allow to make a better profile conversion to the output profile if that's not yet done, but once that's done it is usually fine to send 8-bit/channel data to the printer.

Cheers,
Bart

[digitaldog]

I've not seen any differences. It's only an option printing from a Mac too. If you've been provided 16-bit files, go ahead and print them, no harm and probably faster than sampling down which is pointless.

[John Caldwell]

In my hands, yes there was a difference: 16 bit would crash the driver or cause the paper to spit out huge sheets of un-printed paper, where 8 bit spools would not. This problem arose with physically large files, large print dimensions at large dpi, and 16 bit. Once I found 16 bit selection in the driver to be the cause, I have never again printed in 16 bit with no regrets.

John-

[Schewe]

The 16-bits will allow to make a better profile conversion to the output profile if that's not yet done, but once that's done it is usually fine to send 8-bit/channel data to the printer.

I agree...and this would be important if the color space being used is ProPhoto RGB. The best bet would be to do a color space transform from PP RGB to the paper profile and then drop the mode to 8 bit.

[jrsforums]

I, sort of, agree.  However, you are then "stuck" with that paper/printer combination.

Bart and I discussed this sometime back.  A better concept is to use a "printRGB".  That is a profile smaller than PPRGB, but large enough to include all the possible profile areas of specific paper profiles.  It is then converted to 8bit and can be safely used with any paper profile.

The one I use was developed by Mike Chaney (Qimage) and is called dRGB.icc. I believe Bart or others mentioned similar profiles.

John

[Wayne Fox]

I agree...and this would be important if the color space being used is ProPhoto RGB. The best bet would be to do a color space transform from PP RGB to the paper profile and then drop the mode to 8 bit.

curious why you would need to use this route.  Why not just leave the 16bit checkbox in the print dialog unticked and let LR or PS go ahead and do it’s thing.  I’ve never seen an issue with this.

[Pete Berry]

After experiencing some sky gradation problems several years ago I've been routinely exporting 16-bit PP-RGB's from ACR, and printing through my iPF5100's excellent 16-bit  PS plugin. I've heard it's not true 16-bit output, but even 10-bit gives 4X the gradation steps.

I'm wondering if converting old 8-bit aRGB TIFFs to 16-bit PP-RGB gives any theoretical advantages. On the surface it would seem not once an 8-bit smaller color space is baked-in...

Pete

[jrsforums]

curious why you would need to use this route.  Why not just leave the 16bit checkbox in the print dialog unticked and let LR or PS go ahead and do it’s thing.  I’ve never seen an issue with this.

I print via Qimage for the superior interpolation and smart sharpening....and flexibility and ease of printing.

[Ernst Dinkla]

Larger color spaces and 16 bit profile conversions is one thing. I also wonder whether a 16 bit color image is actually not going through a bottleneck when the the data is split up on the CMYKetc ink channels of the printer. I have a gut feeling that there could be a quality improvement when a 16 bit grayscale image is split up on 6 or more monochrome ink channels (custom solutions), each channel covering part of the tone range and by that no bottleneck on the information transfer. I do not have a printer like that.

Met vriendelijke groet, Ernst

http://www.pigment-print.com/spectralplots/spectrumviz_1.htm
December 2014 update, 700+ inkjet media white spectral plots

[Sbarroso]

I only have (little) experience with my Canon Pixma 1. Here 16 bit printing is only available trough the Canon plug in and selecting there the XPS driver, not directly in Lightroom nor Photoshop (even if the XPS driver is selected).

I did notice that very smooth B&W gradations (i.e. skies) show banding if printed in 8 bits. Using the plug-in+XPS driver (=16 bits) the print shows no banding. I did not noticed banding (yet) in color pictures. I think I did print at least one color picture with a large portion of sky in 8 bits without noticing banding.

Cheers
Santiago

[Wayne Fox]

I print via Qimage for the superior interpolation and smart sharpening....and flexibility and ease of printing.

That’s nice.  I was curious about Jeff’s opinion because I’ve been very pleased with a Lightroom workflow including output sharpening and this seems to be slightly contrary (although I do leave the 16bit checkbox ticked unless I’m printing a pretty large print).

 For me printing from LR is simple, high quality, easily repeatable, no extra software or steps. Maybe I  could squeeze a “little” bit more out of the printer with a RIP, but I even uninstalled the RIP from my 11880 here in my store because I really couldn’t see enough difference to make it worthwhile to make the extra effort worthwhile.

[Schewe]

That’s nice.  I was curious about Jeff’s opinion because I’ve been very pleased with a Lightroom workflow including output sharpening and this seems to be slightly contrary (although I do leave the 16bit checkbox ticked unless I’m printing a pretty large print).

My opinion is that Qimage is a nice app for printing from Windows. Since I don't do Windows, I won't be using Qimage for anything other than testing and doing screen shots in Windows for my printing book. But all that aside, I'm quite happing printing out of Lightroom and even if they made a Mac version (which I'm pretty sure Mike would never do) I would still be using Lightroom for printing.

[jrsforums]

My opinion is that Qimage is a nice app for printing from Windows. Since I don't do Windows, I won't be using Qimage for anything other than testing and doing screen shots in Windows for my printing book. But all that aside, I'm quite happing printing out of Lightroom and even if they made a Mac version (which I'm pretty sure Mike would never do) I would still be using Lightroom for printing.

I know many people that are quite happy printing from Lightroom.  I know no one who has done a side by side test of LR and Qimage who does not think that Qimage does a better job.  Qimage's interpolation and sharpening routines just produce higher quality images.

It is a bit like the discussion on the ppi to send printers.  For years, the "experts" pooh-poohed Mike's recommendation of the need to send images to the printers with the ppi they expected, e.g. 360/720 for Epson.  Finally, after years, the experts did some testing of their own and now the 'tablets from the mountain' now recommend what Mike had recommended all along.  :-)

[JRSmit]

I did try qimage and mirage also. I am very pleased with lightroom for printing. Different is not always better.
i do print at 720ppi and fine details on, actually always. The print time impact is on the whole insignificant. Customer response is!
i am on Windows 7 by the way.