The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

MFD File 16bit to 8bit for Lab

David Schneider

New member
If my lab wants jpg files, which are 8 bit, to do the printing and I'm using a medium format camera that produces 16bit files, what if anything am I losing when there's the conversion from 16bit to 8bit for printing?

Is there a way to maximize the advantages of 16bit before converting to 8bit jpg?
 

Ben Rubinstein

Active member
To be honest you don't lose anything, heck very few inkjets print in 8 bit. 16 bit holds the information needed to make changes, the lab won't be doing any so it should be a case of WYSIWYG.
 

Shashin

Well-known member
+1. Don't worry about it. There really is no loss.

BTW, if you want 16-bit JPEGs, you can save as JPEG 2000. But not everything can read the format.
 

Jack

Sr. Administrator
Staff member
Actually you do loose quite a bit IMHO: First you do loose subtle hue transitions, like in a long sky gradient. When you loose the interim 16-bit colors, you can get posterization or banding in these smooth tones. If you've ever seen a sunset image posted where it looked like rings around the brightest areas, that's due to bit compressions. Anyway, to alleviate this, most good printers will dither it out so it is not so apparent. Second, you loose the more highly saturated colors you can get from most contemporary printers. We show this all the time on our workshops -- identical prints printed from a lab that "desires" srgb jpegs, compared to prints off Epson printers -- the difference is quite visible and significant when the prints are laid side-by-side.
 

Shashin

Well-known member
Jack, can you show examples. I have never had a problem when an image is properly processed and then simply down sampled to 8-bits within the same color space.
 

David Schneider

New member
Actually you do loose quite a bit IMHO: First you do loose subtle hue transitions, like in a long sky gradient. When you loose the interim 16-bit colors, you can get posterization or banding in these smooth tones. If you've ever seen a sunset image posted where it looked like rings around the brightest areas, that's due to bit compressions. Anyway, to alleviate this, most good printers will dither it out so it is not so apparent. Second, you loose the more highly saturated colors you can get from most contemporary printers. We show this all the time on our workshops -- identical prints printed from a lab that "desires" srgb jpegs, compared to prints off Epson printers -- the difference is quite visible and significant when the prints are laid side-by-side.
Jack,

I've been carrying on an email conversation about this with my lab today. I copied your post and asked him about it. This is his reply (and I asked his permission to post any of his comments):

David,

Epson injet and Photographic papers to handle things a lot differently. Yes, those problems can happen with 8 bit images, but again, we don't have a way to handle, injest, or print an image that is not 8 bit.

In a custom print environment, the printer can do more "hand holding" to the file as they are sent to the printer. We don't have that type of production system.
 

Jack

Sr. Administrator
Staff member
Jack, can you show examples. I have never had a problem when an image is properly processed and then simply down sampled to 8-bits within the same color space.
There is absolutely NO WAY to "show" examples online since the web is sRGB! It is why we do show it actual prints on our workshops :)

However, what you can do is download our printer evaluation image, load it into Photoshop running on a computer that has a wide gamut monitor, and then soft-proof it using a profile for Epson Exhibition Fiber on an x900 printer and compare it to the sRGB soft-proof -- the difference in color range will be quite apparent. You will see it in a big way in high greens and to a lesser extent in the reds and blues, but still obvious.

Printer evaluation image can be found on Uwe Steinmuller's site via a link here as well as a link to an explanation of how I use it: http://www.outbackprint.com/printinginsights/pi048/essay.html
 

Jack

Sr. Administrator
Staff member
Jack, how do you guys send prints to the Epson? Give an procedural example ... pretty please :)

-Marc
Easy. First, I convert my MF files to either 16-bit Profoto or 16-bit camera space in C1. Then once in PS, tweak image with localized edits as desired. Save as layered working copy. Size to appropriate output size for Epson printer, which is usually whatever size print I want at 360 PPI. Sharpen for that output size. Soft proof for desired output media and globally tweak contrast via Curve and HSB as/if needed. I flatten and save as imagename_PRINT_size, either as a tiff or psd/b.

Now I hit cmd-P. In the Epson printer dialog, there is an option to use 16-bit data and I do check that and use it. I select the proper settings and profile for the media I am using (my normal ones are all saved as output presets in the Epson dialog dropdown) and hit okay. You are now done and can go to the next image or watch that one come out of the printer.
 

Jack

Sr. Administrator
Staff member
Jack,

I've been carrying on an email conversation about this with my lab today. I copied your post and asked him about it. This is his reply (and I asked his permission to post any of his comments):

David,

Epson injet and Photographic papers to handle things a lot differently. Yes, those problems can happen with 8 bit images, but again, we don't have a way to handle, injest, or print an image that is not 8 bit.

In a custom print environment, the printer can do more "hand holding" to the file as they are sent to the printer. We don't have that type of production system.
At least he's being honest -- and to me that says a lot! The lab you want to run away from is the one that tells you it won't or doesn't matter :)
 

David Schneider

New member
Jack,

I was wondering, are there any labs that do print from 16bit files?

My lab is H&H for most of my work and they are pretty upfront with what they can and can not do. It does help a tiny bit that I photograph my customer service person's family portraits when she makes the trip from Kansas City to NJ.
 

Jack

Sr. Administrator
Staff member
David, send a PM to Wayne Fox on this forum -- he has a commercial lab and Epson printers. Guy and I will do one off custom printing on our Epsons as well, but do not make a big deal out of it.
 

Shashin

Well-known member
However, what you can do is download our printer evaluation image, load it into Photoshop running on a computer that has a wide gamut monitor, and then soft-proof it using a profile for Epson Exhibition Fiber on an x900 printer and compare it to the sRGB soft-proof -- the difference in color range will be quite apparent. You will see it in a big way in high greens and to a lesser extent in the reds and blues, but still obvious.
But aren't you just taking a large gamut image and converting it to a smaller gamut sRGB? That does not show a bit depth problem, but a color space one where you are simply trying to plot what can't be represented in one color space from another.
 

Bill Caulfeild-Browne

Well-known member
I generally print using ProPhoto color space and Canon's 16 bit PS Print Driver to an iPF printer. I have occasionally printed using JPG. The results are inferior both in terms of tonality and gamut, though it's more obvious with subtle subjects than brightly saturated shots..

Bill
 

Jack

Sr. Administrator
Staff member
But aren't you just taking a large gamut image and converting it to a smaller gamut sRGB? That does not show a bit depth problem, but a color space one where you are simply trying to plot what can't be represented in one color space from another.
No, it's definitely a bit-depth issue as well.
 

fotografz

Well-known member
Easy. First, I convert my MF files to either 16-bit Profoto or 16-bit camera space in C1. Then once in PS, tweak image with localized edits as desired. Save as layered working copy. Size to appropriate output size for Epson printer, which is usually whatever size print I want at 360 PPI. Sharpen for that output size. Soft proof for desired output media and globally tweak contrast via Curve and HSB as/if needed. I flatten and save as imagename_PRINT_size, either as a tiff or psd/b.

Now I hit cmd-P. In the Epson printer dialog, there is an option to use 16-bit data and I do check that and use it. I select the proper settings and profile for the media I am using (my normal ones are all saved as output presets in the Epson dialog dropdown) and hit okay. You are now done and can go to the next image or watch that one come out of the printer.
OMG! Jack ... thank you so much!

For lack of checking one little box in the printer dialog, I've missed part of the abilities I've paid for with these cameras.

Folks, listen to Jack ... I've seen some of his big prints first hand, and they are amazing.

-Marc
 

David Schneider

New member
Marc,

I'm kind of happy I asked a question I thought was pretty basic. Looks like we both got something out of the discussion.


I was recommended to Redipix.com. Looks like they print on HP Z3100 and prefer ProPhoto RGB for 16 bit files. Interesting how they can partial mount a gallery wrap and you finish it off; saves on shipping.
 

Craig Stocks

Well-known member
I've learned the hard way that I need to be careful with color space conversions, especially in sRGB. The problem here is the size of the color space, which I find to be the larger problem when converting to sRGB. Inkjet print profiles tend to be much larger color spaces and are less problematic.

Lightroom (and I suspect Capture One as well) can only do so much to squeeze a large color space image into an sRGB output. In LR, the histogram displays distribution and clipping in the native, very large color space. I've had a number of images that look great in LR, but when exported to a JPEG in sRGB some colors start to clip, and the detail in those areas gets blocked up.

Photoshop behaves exactly the same way, converting from ProPhoto to sRGB for instance. At least in PS, you can use soft proofing and the gamut warning overlay to preview out of gamut colors.

I recently posted some examples at Craig Stocks Arts Tutorials - Managing Color Spaces in Lightrom and Photoshop illustrating the problem.
 

Guy Mancuso

Administrator, Instructor
GetDPI Printing workshop and shoot coming in a few months attend it folks. LOL

BTW here in Phoenix at my place.
 
Top