Pages: 1 [2]   Go Down

Author Topic: large prints from Canon 5DII file  (Read 1601 times)

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: large prints from Canon 5DII file
« Reply #20 on: December 09, 2018, 07:45:56 pm »

If you are using an Epson printer ( you didn’t say) yes 180 ppi seems to be the sweet spot with good output sharpening added. But if you are on an HP or Canon I would suggest 150 before beginning to fake pixels. I’ve done that for 12 years with the thermal printers with careful sharpening and or noise reduction.

Bruce Fraser also said in Real World Camera Raw, - I have only one thing to say about interpolation, don’t do it................

John

John, my understanding is that there's a subtle factor here we should be aware of. Taking Epson printers for example - the generally used native input resolution to the printhead is 360 PPI. So if we send 180 to the printer, somewhere along the print pipeline, that 180 gets resampled on the fly to 360 for printing. Hence I believe when we say not to send less than 180, what we're really saying is that this resampling process becomes ever more obviously sub-optimal the further below 180 we're calling upon the system to resample from; and in general, the less resampling the better, no matter how it's done, especially if the prints will be viewed close-up.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."

deanwork

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 2400
Re: large prints from Canon 5DII file
« Reply #21 on: December 11, 2018, 08:57:44 am »

Right. I was referring to manual resampling before the file is sent to the printer, not the resampling done in the printer driver, which is always going on behind the scenes one way or another.


John, my understanding is that there's a subtle factor here we should be aware of. Taking Epson printers for example - the generally used native input resolution to the printhead is 360 PPI. So if we send 180 to the printer, somewhere along the print pipeline, that 180 gets resampled on the fly to 360 for printing. Hence I believe when we say not to send less than 180, what we're really saying is that this resampling process becomes ever more obviously sub-optimal the further below 180 we're calling upon the system to resample from; and in general, the less resampling the better, no matter how it's done, especially if the prints will be viewed close-up.
Logged

Mark D Segal

  • Contributor
  • Sr. Member
  • *
  • Offline Offline
  • Posts: 12512
    • http://www.markdsegal.com
Re: large prints from Canon 5DII file
« Reply #22 on: December 11, 2018, 09:17:43 am »

Right. I was referring to manual resampling before the file is sent to the printer, not the resampling done in the printer driver, which is always going on behind the scenes one way or another.

Understood. But perhaps worth asking whether the word "always" in your statement above always applies - in the sense that if Lr or Ps for example are sending 360 PPI data to an Epson printer, there would appear to be no need for any further resampling to happen under the hood. So if we're not starting with 360PPI data, then the question becomes whether the higher quality resampling happens in Lr/Ps, or elsewhere in the print pipeline. It's kind of useful to know this for optimizing print quality. I would not send 180PPI data to the printer knowing that a sub-optimal process is going to resample it to 360 for making the print; rather, I would prefer to use the higher quality resampling tools for doing this before sending the data to the printer.
Logged
Mark D Segal (formerly MarkDS)
Author: "Scanning Workflows with SilverFast 8....."
Pages: 1 [2]   Go Up