Canon Imageprograf Pro-1000 input resolution & 16bit printing

[dgmkd]

I recently acquired a Imageprograf Pro-1000, my first decent inkjetprinter after years of traditional darkroom printing. Right now I'm trying to wrap my head around some of the detailed technical aspects.

• First, it's been mentioned on this website that Canon printers work with a 300-600 PPI native resolution and that image in other PPI settings are resized 'under the hood'. Therefore it's 'best' to resample to these resolutions to have ultimate control over sharpening. Some press releases also mentioned that the Pro-1000 accepts up to 1200 PPI input files, but I can't find anything about this in the printer specifications. (See http://www.canon-europe.com/about_us/press_centre/press_releases/consumer_news/printers/ipf_pro_1000.aspx) Does that mean that the best way to send files to the printer is now 300-600-1200 PPI? I do have some files that are around 1050 PPI for output size.
• Second, is it possible to do 16bit printing out of Photoshop? I'm currently running CS5 64bit. Is this done automatically when using the XPS driver? I'm slightly baffled that Canon still uses the 32bit Adobe CMM for black point compensation in Print Studio Pro. Printing on matte papers out of PSP is useless without BPC.

I already contacted Canon with these questions and I'm majorly dissapointed by their support. The Canon tech I communicated with, didn't even know the difference between ppi and dpi!

[Bart_van_der_Wolf]

I recently acquired a Imageprograf Pro-1000, my first decent inkjetprinter after years of traditional darkroom printing. Right now I'm trying to wrap my head around some of the detailed technical aspects.

• First, it's been mentioned on this website that Canon printers work with a 300-600 PPI native resolution and that image in other PPI settings are resized 'under the hood'. Therefore it's 'best' to resample to these resolutions to have ultimate control over sharpening. Some press releases also mentioned that the Pro-1000 accepts up to 1200 PPI input files, but I can't find anything about this in the printer specifications. (See http://www.canon-europe.com/about_us/press_centre/press_releases/consumer_news/printers/ipf_pro_1000.aspx)

Hi,

They indeed say that "To enhance quality even further, the printer also accepts 1200ppi high resolution image data, a result of the increasing number of pixels of digital cameras in the market today.".

That does seem to indicate that the data is not simply down-sampled by the printer driver, but it is used in some way. I remember an older HP printer that I used to have, that also allowed to effectively use 1200 PPI. This was confirmed by Qimage, the program I use for printed output, which shows the feedback from interrogating the printer driver as to what it expects and how large the physical print area really is. Since Canon makes specific mention of this, it may actually be a new feature (although the specifications mention "2400 x 1200 dpi", and DPI is not PPI ).

Does that mean that the best way to send files to the printer is now 300-600-1200 PPI? I do have some files that are around 1050 PPI for output size.

I does sound like it should make a difference. I would in any case not down sample to 600 PPI if you have more, but upsample to 1200 PPI instead. It of course depends on image content whether the difference is visible, but upsampling and output sharpening at a larger size offers more control over output sharpness quality. And who knows, maybe the printer also uses a different dithering with smaller droplets. It also means that printhead alignment needs to be more precise.

For those with the printer, it may be useful to try comparing a print of my Printer/Media Resolution test target, of course on high resolution glossy paper. Since the target is designed for a maximum of 600 PPI, printing it at 1200 PPI would result in half the size, and double the resolution numbers as marked. So if the printer cannot resolve detail between 300 and  600PPI, which becomes 600 to 1200 PPI at half the size, then maybe only the dithering is finer. But maybe it can really resolve such fine detail.

• Second, is it possible to do 16bit printing out of Photoshop? I'm currently running CS5 64bit. Is this done automatically when using the XPS driver? I'm slightly baffled that Canon still uses the 32bit Adobe CMM for black point compensation in Print Studio Pro. Printing on matte papers out of PSP is useless without BPC.

That may depend on the operating system. Under Windows it should be possible to print at 16 bit/channel when using the XPS driver. How that interacts with CS5, I do not know.

I already contacted Canon with these questions and I'm majorly dissapointed by their support. The Canon tech I communicated with, didn't even know the difference between ppi and dpi!

Ah well, there are more who do not understand the distinction.

Cheers,
Bart

[Mousecop]

I've had my Pro-1000 for a few weeks, and just started printing on 17x22 paper. I'm printing using Lightroom 6, images captured using an Olympus E-M1, which generates a 16mp 14-bit image. So far, I'm printing at 600ppi.

The prints look outstanding. I don't see any pixellation at all, even with a magnifying glass. I would literally need a microscope to see the ink dots. I might test 1200ppi printing, but at the moment I don't think the printer will use the extra data.

I assume you're capturing at a higher resolution than I am. Even so, I don't think there will be a perceptible difference between 600ppi and 1200ppi at the largest paper size, at normal viewing distances (24" - 30").

In terms of bit depth, I know LR supports 16 bit, so I assume PS does as well. For Windows, it's available via the XPS driver. I don't have access to my system right now, but you might need to check a box, and/or go through Print Studio Pro.

I'm not too surprised that a Canon tech isn't trained on this, as most of their requests are probably about how to uninstall the driver and swap inks. They might get better over time. Fingers crossed.

[Bart_van_der_Wolf]

I've had my Pro-1000 for a few weeks, and just started printing on 17x22 paper. I'm printing using Lightroom 6, images captured using an Olympus E-M1, which generates a 16mp 14-bit image. So far, I'm printing at 600ppi.

The prints look outstanding. I don't see any pixellation at all, even with a magnifying glass. I would literally need a microscope to see the ink dots. I might test 1200ppi printing, but at the moment I don't think the printer will use the extra data.

Hi,

Is there a specific setting in the printer driver for 1200 PPI?

The document on their websitewebsite says:

Crystal-Fidelity
The new imagePROGRAF PRO allows EOS camera users to faithfully print their photos,
retaining the details and clarity of their original images captured using Canon EOS cameras.
This high level print quality is only achievable by having a Canon input-output solution, using our dedicated photo editing software – Digital Photo Profressional and the Print Studio Pro plug-in.

I assume you're capturing at a higher resolution than I am. Even so, I don't think there will be a perceptible difference between 600ppi and 1200ppi at the largest paper size, at normal viewing distances (24" - 30").

I do not think that for average images the difference's above 600 PPI will be too obvious, but it does allow much better sharpening precision, and in case there is Vernier acuity potential (very tiny abrupt displacements) in the image. Additionally the dithering may be different.

In terms of bit depth, I know LR supports 16 bit, so I assume PS does as well. For Windows, it's available via the XPS driver. I don't have access to my system right now, but you might need to check a box, and/or go through Print Studio Pro.

Yes, but on Macs the print pipeline is partly handled by the operating system, and the last thing I've read suggests that that prohibits 16-b/ch output.

Cheers,
Bart

[henrikolsen]

For those with the printer, it may be useful to try comparing a print of my Printer/Media Resolution test target, of course on high resolution glossy paper. Since the target is designed for a maximum of 600 PPI, printing it at 1200 PPI would result in half the size, and double the resolution numbers as marked. So if the printer cannot resolve detail between 300 and  600PPI, which becomes 600 to 1200 PPI at half the size, then maybe only the dithering is finer. But maybe it can really resolve such fine detail.

I have the printer, and have tried printing your target now at native size, 600 dpi, from Photoshop CC 2017 on a Canson HighGloss RC paper (highest quality setting, and media type as recommended by Canson). I'd like to do the comparison mentioned (1200 dpi), and report back. But how to print at 1200 dpi? I cannot set resolution (as I know of) in printer driver, so how do I best scale your target before/at print to ensure it stays as intended?

[Bart_van_der_Wolf]

I have the printer, and have tried printing your target now at native size, 600 dpi, from Photoshop CC 2017 on a Canson HighGloss RC paper (highest quality setting, and media type as recommended by Canson). I'd like to do the comparison mentioned (1200 dpi), and report back. But how to print at 1200 dpi? I cannot set resolution (as I know of) in printer driver, so how do I best scale your target before/at print to ensure it stays as intended?

Hi Henrik,

Try printing at half of the original 130 mm square size, so at 65 x 65 mm (if resampling can be disabled). Alternatively, change the PPI setting of the file by resaving a copy of the target after resizing without resampling by changing the PPI to 1200PPI and print at original size (now 65 x 65 mm). One can only print at 1200 PPI if the detail is there (more than 600 PPI of detail at the requested output size).

Cheers,
Bart

[Geraldo Garcia]

To print using the 1200ppi you cannot use de xps driver (it is limited to 600ppi, don't ask me why). You also need to disable "Prevention of print data loss" on the "print options" under the "page setup" tab of the driver's window.

[Bart_van_der_Wolf]

To print using the 1200ppi you cannot use de xps driver (it is limited to 600ppi, don't ask me why). You also need to disable "Prevention of print data loss" on the "print options" under the "page setup" tab of the driver's window.

Good points, thanks for adding that.

The "Prevention of print data loss" should only be activated if large format AND high-resolution prints are made when they lead to print buffer overflow and memory issues. It is a destructive setting that alters/compresses (lossy ) the data.

Why the XPS driver is limited to 600PPI is a riddle indeed

Cheers,
Bart.

[henrikolsen]

To print using the 1200ppi you cannot use de xps driver (it is limited to 600ppi, don't ask me why). You also need to disable "Prevention of print data loss" on the "print options" under the "page setup" tab of the driver's window.

I'm on macOS. Don't know if anything special applies here...

[henrikolsen]

Try printing at half of the original 130 mm square size, so at 65 x 65 mm (if resampling can be disabled). Alternatively, change the PPI setting of the file by resaving a copy of the target after resizing without resampling by changing the PPI to 1200PPI and print at original size (now 65 x 65 mm). One can only print at 1200 PPI if the detail is there (more than 600 PPI of detail at the requested output size).

Picked up some more glossy paper now for further testing. Did an auto head alignment first for this test paper, swallowing three sheets, and now the sheets are flying, testing settings

One set of tests has revealed something interesting. Normally I print with application managed colors, picking a suitable ICC, highest driver quality - and now unidirectional, as I see an improvement on the formal tests here. The 1200 dpi (changed PPI in Photoshop to 1200 dpi without resampling) with these settings did not resolve better, it seems.
BUT! Looking at the small clarkvision graphic mentioned earlier (very easy to evaluate with clear pixel boundaries, e.g. one 1px line black, next one white, then black (not grays or other aliasing in between), and choosing "Printer manages color" _and_ "Canon color matching" (not Colorsync), I can see the print resolve the 1px thin horizontal lines at 1200 dpi.
But color management is down the drain then I guess, as I wouldn't know how to control it, choosing ICC, rendering intent etc. Perhaps a chosen custom media type from MCT can have an attached ICC respected, but with what intent then?

Using the setting on 600 dpi prints also improve results. I notice less banding/waves/aliasing on your target.

Can anyone explain, or check with Canon, why this combination seems to be the only one triggering a higher resolution ability in the printer/driver?

Any way to get this same sharpness/resolution with color management?

I've seen this sharpness/resolution increasement before when testing driver settings, but on photographs, not resolution tests. Not obvious on photos, but enough that I noticed, where I expected no difference (other than to see how colors where handled by Canon driver itself with me choosing an ICC profile). Interesting to see it confirmed now on more formal tests.

Using macOS. Don't know how similar tests would work out in Windows.
I generally cannot get vertical line separation as distinct as horizontal, even after new auto head align. Might try a manual one, but I assume the automatic should be hard to beat (experiences here?). Perhaps this behaviour is normal. It might seem as if vertical lines are made of stacked small slanted lines of dots, instead of a straight and aligned set of vertical dots - the horizontal line is straight, not pieced together of small slanted lines.

[henrikolsen]

Hmmm. Seems to be some inconsistency regarding 1200 dpi functionality. Today I had 1200 dpi working for a time (seeing 1px horizontal line separations at 1200 dpi), both with and without use of application color management (yesterday I only saw it working for Printer manages color). Then suddenly it wouldn't work for either setting (printing gray lines where the 1px black and white lines should have been separated, perhaps as if downsampled and averaged to 600 dpi - don't know, except it didn't resolve anymore. Power cycle of printer and OS, and repeating the tests, it now resolves at 1200 dpi again - with same settings tried before. Strange, also considering how it sometimes slows down immensely.

[Geraldo Garcia]

Sometimes we can get a bit carried away with this kind os specs. It is worth to remember:

1) The printer will always print at its maximum resolution of 2400x1200 DPI. Although the speed of the printing and options like "unidirectional printing" may improve the precision and the perceived detail, they do not change the number of DPI.
2) The driver settings will affect the expected image resolution in PPI to be fed to the printer as the dithering algorithm may automatically downscale images that exceed the expected PPI resolution. With the highest settings you may feed 1200PPI images, with the lower settings the printer expects images with a resolution of 600PPI or less.

Regards.

[henrikolsen]

New details now. The media type chosen has an effect on the end result resolving power. How exactly the relationships are I don't know yet, but would love to know to make some better informed decisions setting up for best quality.

Feeding the printer driver with 1200 PPI test image, I can now confirm it is accepted and utilised using Canon Photo Paper Plus Glossy II media type (matching my test paper here, not waisting more expensive media on this) and using "Highest" Print Quality. Haven't tried lower quality settings, yet.
Edit: Tried at "Standard" quality, and downsampling is happening, guessing to 600 PPI - matching Garcia's post.
Changing media type to a CO OFF variant, again on "Highest" the 1200 PPI fed is _not_ utilised anymore and resolution clearly downsampled, 600 PPI still resolved though. Media type selection is causing this.

A more surprising observation is using a custom media made in MCT (Canons Media Configuration Tool) based on the (by the tool given weight/thickness) recommended HeavyPhotoPaperA, feed adjustment done and ink level and head height set as OEM media type (medium-small and low). Feeding it 1200 PPI on "Highest" does _not_ utilise the resolution given, again downsampling to, I'm guessing, 600 PPI. This is a surprise to me.

As this 1200 PPI quest is more nerd that practice, I might not care - although still wonder why this has effect or not. Some of Barts demanding resolution tests breaks down anyhow at high frequency patterns at 1200 PPI (perhaps dithering gives up somehow, I see uneven patterns that is smooth/even/sharp at close at 600 PPI).
BUT - what's worse is than even feeding it 600 PPI, the results using the OEM matching media type is still better, and the custom media type inferior. And it does not require a loupe, microscope or test target to see, it's by naked eye (on a snippet of tree detail included with my target - the target is for enlargement, real world image for naked eye tests). Both media types resolve 600 PPI, but the OEM type is quite a lot more refined/sharper. If it's dithering being different I cannot decipher yet, but different the overall result clearly is.

So... As I don't need the 1200 PPI resolvement power, I _do_ want my 600 PPI to look as nice as possible, and observing I cannot achieve this with custom media types (which I setup for 3rd party papers, for calibration, feed adjustment and accounting) has left me disappointed.

Hope someone can clarify / fix. It would be nice with an overview like

 - What is required to have driver/printer utilise 1200 PPI input? (exact driver and media type settings making this possible)
 - How can I make my custom media types match OEM media types in sharpness (not requiring 1200 PPI support, but wanting 600 PPI to look as good as with OEM media type)?
 - Which media types support "Contrast Reproduction" in Print Studio Pro (including base types for custom media)?

Printing from Photoshop CC 2017 running on macOS Sierra.

[henrikolsen]

Can anyone confirm whether head alignments are stored per media type?

I do select a media type when doing alignment, but previously I had a feeling it was a general adjustment spanning all media settings, but from https://support.usa.canon.com/kb/index?page=content&id=ART165903 it says "After adding custom paper, we recommend adjusting the position of the printhead for the custom paper that you added.". I'm wondering if they meant feed adjustment, which is per media type in MCT, and not head alignment...
Or is the alignment actually stored per media type? If so, does it work like calibration fall backs "for uniques", as in general calibration or head alignment applies to specifics (uniques or custom media types) until they optionally get their own specifically done.

If indeed head alignment is per media type, there might be hope I need my custom media type specifically head aligned to have it match the OEM media type which did have a head align (performing way better).

Also, best head alignment (maybe even feed adjustment) should be related to head hight as well (relative to paper surface), but I'm not sure that's how Canon stores it. Anyone?

[henrikolsen]

More testing now reveals the following surprise about (custom) media types.

I discovered today in the accounting manager that the Print Quality wasn't at the same setting when I tried to do equal comparisons between the OEM media type and a custom one. Digging deeper it turns out that "highest" doesn't mean the same on the two media types!
The scale (as reported in accounting manager) goes from 5 to 1, 5 meaning fastest/worst, 1 meaning slowest/best. I've never seen 1 used, but 2-5 are in use. Would like to hear how to get use of setting 1.

Print driver lets you either use a named quality setting or set a custom one (on a slider from fast to fine, covering 5-1, not all being available dependent on media type). On the OEM media I can choose High and Highest, seen in accounting manager to map to 3 and 2 respectively. The custom media type has choices of Standard, High and Highest. I've tested both on Highest yesterday, where I noticed the unexpected quality difference between them. Here comes the strange thing. You'd be expected Highest to mean, well, highest quality - hard to get higher that highest you'd think, but no... If you go to Custom it turns out there are four options available and not just the three named ones. And the mapping of the named ones are offset one toward the faster end, meaning Standard maps to 5, High to 4, Highest to 3. Notice the one-off difference? Highest was 2 before on OEM (better). But using Custom slider, you can drag it one notch higher, matching setting 2 of the OEM. That isn't exactly obvious UI for me.
Can someone inform Canon of this strange inconsistency, or direct me in a suitable direction to do so directly?

Now quality is a better match. And guess what, now that the custom media can be custom set to 2 (one step higher that "Highest" maps to) it also will accept and utilise 1200 PPI input like the OEM at Highest (here mapping directly to quality 2). Only Highest (or rather setting 2 to be precise, as mapping is screwed) utilise 1200 PPI.

All is not solved though, as I still see a sharpness increase using the OEM media type, both testing 600 and 1200 PPI input, even though quality settings behind the scenes seem to match now (level 2). Custom media type has head height and ink level matching OEM comparison (low and medium-small). OEM has better separation of details, hard to describe without having prints at hand, but it's clearly doing something different, and better. Both resolve 600 and 1200 PPI line tests, but with a different perceived quality. Much prefer the OEM. If it's better dithering, up sampling, sharpening or what, I don't know, but better it is. And yes, it can be seen by naked eye, and I dislike having this quality drop using custom media, as I find several good uses of custom media types including Accounting, custom ink level, head height and feed adjustment - even calibration. So it still needs clarification why sharpness is better on OEM media type. Can anyone investigate or explain further?

Later going to check options in Print Studio Pro, I could more easily confirm the quality mapping inconsistencies in it's UI. Rest was done from Photoshop CC 2017 on macOS Sierra, driver version 16.10.1.0.

[Bart_van_der_Wolf]

More testing now reveals the following surprise about (custom) media types.
...

Fascinating. Thanks for sharing your findings so far. Hope that others with the same printer (generation) can add their findings, and I hope to hear more.

Cheers,
Bart

[Sbarroso]

The Canon Pixma Pro-1 cannot print at 16 bits from LR, even using the XPS driver. At least that was the case 2 years ago. I may check again, just in case a update made the miracle.
That was obvious when printing smooth B&W gradients.

There is no problem using the canon plugin, even from LR. XPS driver should be selected.

Things may have changed with the new x000 pro models, but i've no idea.

[GrahamBy]

The custom media type has choices of Standard, High and Highest. I've tested both on Highest yesterday, where I noticed the unexpected quality difference between them. Here comes the strange thing. You'd be expected Highest to mean, well, highest quality - hard to get higher that highest you'd think, but no... If you go to Custom it turns out there are four options available and not just the three named ones. And the mapping of the named ones are offset one toward the faster end, meaning Standard maps to 5, High to 4, Highest to 3. Notice the one-off difference? Highest was 2 before on OEM (better). But using Custom slider, you can drag it one notch higher, matching setting 2 of the OEM. That isn't exactly obvious UI for me.

Is it possible that if you specify "heavy art paper", the printer is expecting something up around 400gsm or 0.4mm, rather than the 310-ish gsm of good gloss/lustre/baryta papers? So far as I'm aware, the only papers in that thickness range are uncoated matte papers, on which the ultimate resolaution might be lost due to ink migration.

If that is the case, it wouln't be an inconsistency, but a sensible choice to offset the quality choices to take into account the capability of the paper.

Canon could certainly do themselves and their customers a favour if they had some of their translations checked by native speakers: aside from the "high" and "highest" nonsense, there is also calling the top paper feed the rear feed, while the rear feed is the manual feed. Plus in PSP2 in French, they've confused verbs and adjectives so that where they mean "activated" they've written "to activate" (phonetically the same, activé or activer, almost opposite meanings). It would seem a small additional investment relative to the total development cost...

[henrikolsen]

Is it possible that if you specify "heavy art paper", the printer is expecting something up around 400gsm or 0.4mm, rather than the 310-ish gsm of good gloss/lustre/baryta papers? So far as I'm aware, the only papers in that thickness range are uncoated matte papers, on which the ultimate resolaution might be lost due to ink migration.

If that is the case, it wouln't be an inconsistency, but a sensible choice to offset the quality choices to take into account the capability of the paper.

I haven't used a "heavy art paper" as basis media type, but HeavyPhotoPaperA as recommended by the configuration tool based on category (photo) and weight/thickness (only one can be entered, but they agree no matter which is entered here). Feed adjustment done and ink level and head height set as OEM media type (medium-small and low). So I would still assume equal print quality was possible.

I will however try out other basis media types though, although I would much prefer suitable technical information than using a lot of time and money on this testing on my own - possibly without ever achieving what I see with OEM, in case Canon, deliberately or not, made custom media types somehow different in behaviour. So any help on precise information on custom media types, their base media choice and related behind-the-scenes interpretations would be very much appreciated.

Canon could certainly do themselves and their customers a favour if they had some of their translations checked by native speakers: aside from the "high" and "highest" nonsense, there is also calling the top paper feed the rear feed, while the rear feed is the manual feed.

Oh yes! The "rear feed" that's not actually the rear feed also got me confused to start with. Great discussions can be had here, like "Oh, we might misunderstand each other. You mean the rear feed at the top? Not the manual feed at the rear?". Wonderful .

[Geraldo Garcia]

The "manual feed" is at the back and the "Rear feed" is at the top. That perfectly summarizes what we face in terms of nomenclature, interface and documentation on Canon printers.