I have never had a 20x30 print made and was wondering do I have to upsize the image so the print size is 20x30? Also, where can I get information about how to blend images? Thanks for your time.
"I have never had a 20x30 print made and was wondering do I have to upsize the image so the print size is 20x30?"
These days, many commercial printers use what's called raster image processor (RIP) software that can do a far better job of resampling an image than does Photoshop or other resizing software.
For instance, I had an image from a 6MP Canon 10D that I wanted enlarged to 20 x 30 and El-Co Color Labs was recommended to me. At $10 a print (2 print minimum, though), I could afford to experiment. After emailing back and forth with them, I sent them an image file at the original size. I was absolutely amazed at the quality of the print I received from them! Have used them many times since and have been extremely pleased with their service.
Needless to say, in order to get a good enlargment, you have to send them a high quality file. Using a tripod, a high quality lens and RAW capture/editing is a good start. Don't get too carried away with the sharpening! Subject matter also has a lot of bearing on the quality. If you're interested in this service and have any questions, ask away...they have great customer service. Be sure to read the FAQ page linked there first.
"Where can I get information about how to blend images?"
No need to submit your image sized at 20x30. Just submit your image in either tiff or jpg format and the lab will print it at your desired size. You just need to make sure your image has enough detail/quality to enlarge that size without getting too pixelated.
I agree also with John and Jon, I have made 20 x 30 prints from a 20D up to 5Dmkii camera, and have always been advised by the lab to send the file at it's native size. The labs do a fantastic job re-scaling the file these days.
Interesting advice here... I've always resized in PS before sending the files to the lab, and I've gotten great results, but I'm curious about sending the native size. My issue would be with sharpening... Do they do any additional sharpening for you? It just seems like the sharpening would be a bit of a crapshoot on my part...
The only time I haven't 'rolled my own' is when I want an enlargement as big as a 20 x 30, when I used the lab mentioned above. Much less of an issue now using a 21MP camera.
Re: sharpening, according to the FAQ from that lab--"Sharpen it to what you like without bringing out artifacts. Although the RIP applies sharpening based on the size of your file and the print size it is printing, it is only enough to make up for what is lost during the normal printing of digital files to photographic paper."
In a nutshell this is what two separate labs have told me over the years
DPI or PPI are relative measurements that can mean many things depending on the application you are preparing your image for. For the purposes of digital photo-finishing this measurement has no meaning and will not effect your image file. For example, an image that is 1024 x 768 at 100dpi is the exact same file as an image that is 1024 x 768 at 500dpi. Both files contain a total of 786432 pixels.
This per MPIX
Our printing equipment will resize the files as needed when they are printed. Like sharpening an image, re-sampling and interpolation is very much dependent upon the particular printer being used and the print size being created. Interpolating an image without knowing the specific characteristics of a printer can be counterproductive. Re-sampling and interpolating images take time, time that is better spent on the creative aspects of your photography
That maybe slightly OT, but basically all the labs are interested in is the information contained in the file, so obviousily the more pixels the higher the quality. This is why there is no need to re-size your original file, except for cropping (without affecting resolution) when needed. Most online labs will list their recommended "optimal resolution" for certain print sizes, i.e. MPIX recommends 5000x7500, but a min. of 2000x3000 for a 20 x 30 print.
If you're creating art for yourself or clients, I think complete creative control is extremely important. I want to get back exactly what I send, and I want it to look as good as possible. Getting a 20x30 from a smaller file (12mp~) printed at a good resolution (300dpi for example, a good number for a bit of viewing distance and retained small detail - so 9000x6000) takes significant post work to get looking great. Letting a lab do this with their RIP and mystery global one-size sharpening does not seem wise to me. Especially with the advice to sharpen at native before they up-res and re-sharpen?! What?!
I do several sharpening steps (global, local, output, etc), sometimes a bit of grain to break up the algorithmic "fake" pixel creation look, etc etc. For example, shadow areas turn to mud without specific local attention with a large up-res. It never fails that I get *exactly* what I sent, back in paper form.
I don't mean to come off like a jerk or anything, and I'm sure many here have gotten wonderful results from this method. And it certainly is true that sending your file at a native resolution IS better if it lands between 200~ and 400~ DPI at desired size (though I'd not go lower than 240 unless extremely large). But I won't be giving up the keys to the castle anytime soon.
I'm of an even really differenter opinion on this. I like to downsample to a 20X30 print, so I shoot bigger film and scan it to the best of my ability.
But if I've got to go the other direction from a small file to a large print, then I like to test before printing because one formula does not fit every solution. If you can, try sending two crops for print at 8X10 (big enough to determine quality). Let one be the native file dimension with their RIP doing the work. Upsize the other and do the sharpening yourself on the other. It helps if you know the native resolution of the printer they'll use and set your enlargement to that resolution. Thats the only way to keep their RIP from resizing your image. Otherwise their RIP is still going to do its mystery work on your image. Then look at these prints side by side to determine which best fits your tastes.
I actually have tested this a bit. I sent a file at camera native resolution, printer native resolution and under camera native resolution, each sharpened to make a 4X6 print. So lets say the printer is a 300ppi printer and I sent them files at 240, 300 and 400ppi. The 300ppi file was much sharper than the other two (too sharp in fact). I did this again on my Epson printer at home with 4X6 prints with 240, 360 and 400. The 360ppi image was much sharper than the other two as well.
But this is different when downrezzing than when uprezzing from the camera's capability and there will probably be some times when the RIP will outperform PS Bicubic and other times where Bicubic wins the day.
The range of comments reveals the best answer to your question ... this is really a lab-dependent issue, Jeff. Simply find out what the printer wants ... which can vary wildly from place to place. RIPs can certainly make life easier, if you can find one that produces consistent results across all machines in their lab. Sometimes, your prints will be run through different machines, depending on print size. But I prefer to use places like WHCC or West Coast, where they apply no sharpening whatsoever. And I always know exactly what I'm getting. They leave the control in the photographer's hands - which still seems ideal to me. I'm not saying I know it's the "best" solution, but it's definitely what I like. No guesswork.
A lot of the bigger labs have "printing for dummies" policies in place - for obvious reasons. Sometimes, the RIPs for a given printer really are that much better than what you could do on your own (like LightJet). But generally speaking, equipment and/or RIP capabilities are often just better than the capabilities of the masses who submit files. (There are obviously some exceptions - as people have noted.) And often, the RIPs are designed to accommodate the needs of the high-volume users like portrait photographers - who really have to dump large numbers of images into their system without any prep work. Regardless, it's good to read the reports here. Sending in native files to large labs hasn't always been a safe bet in past years.
But it's not rocket-science, so I'll recommend doing it yourself with a place that encourages photographer control (WHCC, West Coast, etc.) - unless the printer you talk to KNOWS his RIP will do better than you will. And in that case, enjoy the time you're saving!
Jeff ... a good time to try out a few labs, I'd say.
Kurt: Ditto what John said, I didn't think you came off as a jerk AT ALL! This is exactly the kind of conversation I like best about this forum. Let a thousand flowers (ideas) bloom.
Discussion
13 repliesFor instance, I had an image from a 6MP Canon 10D that I wanted enlarged to 20 x 30 and El-Co Color Labs was recommended to me. At $10 a print (2 print minimum, though), I could afford to experiment. After emailing back and forth with them, I sent them an image file at the original size. I was absolutely amazed at the quality of the print I received from them! Have used them many times since and have been extremely pleased with their service.
Needless to say, in order to get a good enlargment, you have to send them a high quality file. Using a tripod, a high quality lens and RAW capture/editing is a good start. Don't get too carried away with the sharpening! Subject matter also has a lot of bearing on the quality. If you're interested in this service and have any questions, ask away...they have great customer service. Be sure to read the FAQ page linked there first.
I've always thought that David Clapp's exposure blending tutorial was pretty straightforward.
No need to submit your image sized at 20x30. Just submit your image in either tiff or jpg format and the lab will print it at your desired size. You just need to make sure your image has enough detail/quality to enlarge that size without getting too pixelated.
Re: sharpening, according to the FAQ from that lab--"Sharpen it to what you like without bringing out artifacts. Although the RIP applies sharpening based on the size of your file and the print size it is printing, it is only enough to make up for what is lost during the normal printing of digital files to photographic paper."
This per MPIX
That maybe slightly OT, but basically all the labs are interested in is the information contained in the file, so obviousily the more pixels the higher the quality. This is why there is no need to re-size your original file, except for cropping (without affecting resolution) when needed. Most online labs will list their recommended "optimal resolution" for certain print sizes, i.e. MPIX recommends 5000x7500, but a min. of 2000x3000 for a 20 x 30 print.
If you're creating art for yourself or clients, I think complete creative control is extremely important. I want to get back exactly what I send, and I want it to look as good as possible. Getting a 20x30 from a smaller file (12mp~) printed at a good resolution (300dpi for example, a good number for a bit of viewing distance and retained small detail - so 9000x6000) takes significant post work to get looking great. Letting a lab do this with their RIP and mystery global one-size sharpening does not seem wise to me. Especially with the advice to sharpen at native before they up-res and re-sharpen?! What?!
I do several sharpening steps (global, local, output, etc), sometimes a bit of grain to break up the algorithmic "fake" pixel creation look, etc etc. For example, shadow areas turn to mud without specific local attention with a large up-res. It never fails that I get *exactly* what I sent, back in paper form.
I don't mean to come off like a jerk or anything, and I'm sure many here have gotten wonderful results from this method. And it certainly is true that sending your file at a native resolution IS better if it lands between 200~ and 400~ DPI at desired size (though I'd not go lower than 240 unless extremely large). But I won't be giving up the keys to the castle anytime soon.
But if I've got to go the other direction from a small file to a large print, then I like to test before printing because one formula does not fit every solution. If you can, try sending two crops for print at 8X10 (big enough to determine quality). Let one be the native file dimension with their RIP doing the work. Upsize the other and do the sharpening yourself on the other. It helps if you know the native resolution of the printer they'll use and set your enlargement to that resolution. Thats the only way to keep their RIP from resizing your image. Otherwise their RIP is still going to do its mystery work on your image. Then look at these prints side by side to determine which best fits your tastes.
I actually have tested this a bit. I sent a file at camera native resolution, printer native resolution and under camera native resolution, each sharpened to make a 4X6 print. So lets say the printer is a 300ppi printer and I sent them files at 240, 300 and 400ppi. The 300ppi file was much sharper than the other two (too sharp in fact). I did this again on my Epson printer at home with 4X6 prints with 240, 360 and 400. The 360ppi image was much sharper than the other two as well.
But this is different when downrezzing than when uprezzing from the camera's capability and there will probably be some times when the RIP will outperform PS Bicubic and other times where Bicubic wins the day.
That was a very well reasoned statement.
A lot of the bigger labs have "printing for dummies" policies in place - for obvious reasons. Sometimes, the RIPs for a given printer really are that much better than what you could do on your own (like LightJet). But generally speaking, equipment and/or RIP capabilities are often just better than the capabilities of the masses who submit files. (There are obviously some exceptions - as people have noted.) And often, the RIPs are designed to accommodate the needs of the high-volume users like portrait photographers - who really have to dump large numbers of images into their system without any prep work. Regardless, it's good to read the reports here. Sending in native files to large labs hasn't always been a safe bet in past years.
But it's not rocket-science, so I'll recommend doing it yourself with a place that encourages photographer control (WHCC, West Coast, etc.) - unless the printer you talk to KNOWS his RIP will do better than you will. And in that case, enjoy the time you're saving!
Jeff ... a good time to try out a few labs, I'd say.