Canon 44" Pro 4600 versus Epson 44" Epson P9570

[alan a]

See my post directly below.  Probably time to toss my old Canon iPF 8400 and replace it.

Any opinions on the Canon 44" Pro 4600 versus the Epson 44" P9570?

Which clogs the most?  Life of the print heads and cost of replacement? 

Any other features that are nice?  (In the old days Canon has a software driver that included the ability to figure margins for printing canvas to stretching on bars.  A pretty minor feature in the scheme of things, but it was nice.)

Does Epson still have a straight feed for really thick stock?  And Breathing Color used to sell metal that could go through the Epson, but they appear to no longer stock that. 

Canon appears to be about $1,000 less.  I placed an order on Friday so I could use the $800 rebate which brings the price down to $3716.  I think Epson had a rebate but it has since expired.

[Czornyj]

Canon never clogs.
PF-10 print head lasts for well over 30l on average.
After PRO-4600 was upgraded with new archival inks the only reason to buy Epson that I can think of is masochism.

[alan a]

Czornyj,

How many generations old is the iPF 8400 (my current printer) compared to the 4600? 

Four? 

8400 --> 4000 --> 4100 --> 4600 ??

You said "PF-10 print head lasts for well over 30l on average."

Dumb question, what unit of measurement is that?  301 of what?

[JCD]

After all the trouble with Epson the last couple of years I was all set to by the Canon 4600.

Really glad I went for the Epson after using it. I’m blown away with the ease of use, speed but especially all aspects of image quality.

Right now though, the price on the 4600 is really great and they come with the 330 ml inks which is icing on the cake.

I wrote some comment about the 9570 on the Articles section of this forum here. It’s not a tech review just my impressions since working with it. Since they resolved these paper feed issues with the latest media installer, Epson Layout Program and new firmware we’ve had zero issues of any kind. We’ve been cranking out 30x40 prints on one for a couple of months. Although I have three other 44” printers this is really the only printer I need. I think it’s spectacular so far. Hadn’t has one missing nozzle yet and it’s been used only one a week so far.

Quick comparisons:

Both very fast

Even with the gloss enhancer channel on the Canon reports are that on fiber gloss still has issues. If so that’s a real drag.

The dmax on Canon Lucia Pro 2 inks on matte rag papers not nearly as good as Epson.

The dither and fine detail of these thermal Canon and HP heads not nearly as good as the 9570/7570.

The Canon rep told me the inks are all new but the head design isn’t.

Both have good color gamut. I know the Epson is excellent. Haven’t used the Canon.

We still don’t know how “archival” the new Canon Lucia 2 inks are because they haven’t been published by Wilhelm or anyone last time I looked. But we know they are better.

Walker who tested the new 60” Canon with the same inks said the bw is ok . More neutral than they used to be.

My bw prints on the 9570 were spectacular and better than any oem inkset ever seen, by far. I would buy this printer for that reason alone and zero bronzing or gloss issues.

The Canon heads in my experience need to be replaced every year, and they aren’t cheap, but you can do that easily yourself. Roll paper feeding on the new Canon is supposed to be awesome. I’ve had zero issues myself with the 9570.

The Epson heads should last a long time now that this head strike issue is fixed. But they are expensive and one should keep this printer in warranty in my opinion.

John

Canon never clogs.
PF-10 print head lasts for well over 30l on average.
After PRO-4600 was upgraded with new archival inks the only reason to buy Epson that I can think of is masochism.

[alan a]

Many thanks for the two replies.

Hard to know which printer to buy based on the replies.  When I scan through the posts, there appears to be many more posts discussing problems with Epson.  Is that because Epson sells many more printers, or Epson in fact has more problems?

What does an Epson print head for the P9570 cost?  JCD recommended keeping the Epson in warranty for that reason.  My guess is that extended warranties cost at least $1,000 per year of Canon is any guide in that regard. 

[JCD]

I got the five year warranty. Most people do. You save a lot with that otherwise about $1,400 annually I think.
These heads are a lot more complex and much bigger than the previous models. It puts down a lot more ink at a time, like the old Epson 10k from early 2000s , one of the reasons it’s faster .  For bi directional it uses both rows of the nozzles and even faster. These 12 inks are amazing and still more permanent than p9000.

These printers are very evolved electronically, more apps and complex to work on. Both Canon and Epson. They are both heavy as hell too, about twice as heavy as my HPs .

Many thanks for the two replies.

Hard to know which printer to buy based on the replies.  When I scan through the posts, there appears to be many more posts discussing problems with Epson.  Is that because Epson sells many more printers, or Epson in fact has more problems?

What does an Epson print head for the P9570 cost?  JCD recommended keeping the Epson in warranty for that reason.  My guess is that extended warranties cost at least $1,000 per year of Canon is any guide in that regard.

[Czornyj]

Czornyj,

How many generations old is the iPF 8400 (my current printer) compared to the 4600? 

Four? 

8400 --> 4000 --> 4100 --> 4600 ??

You said "PF-10 print head lasts for well over 30l on average."

Dumb question, what unit of measurement is that?  301 of what?

Exactly four, I bought PRO-4000 in 2016 right after introduction (I had iPF8300 before PRO-4000). It's 30 liters of ink, usually closer to 40 liters - PF-10 is much more robust and failure proof than PF-5 in iPF8400.

[Czornyj]

When I scan through the posts, there appears to be many more posts discussing problems with Epson.  Is that because Epson sells many more printers, or Epson in fact has more problems?

What does an Epson print head for the P9570 cost?  JCD recommended keeping the Epson in warranty for that reason.  My guess is that extended warranties cost at least $1,000 per year of Canon is any guide in that regard.

Because Canon PRO series have no problems at all - it's just as simple as that. Last time I changed print head was 5 years ago (11.2018), and it still works well. There were initial production issues with PF-10 back in 2016-2017 so they used to fail after a year or so, but Canon fixed it immediately.

[NAwlins_Contrarian]

The dmax on Canon Lucia Pro 2 inks on matte rag papers not nearly as good as Epson.

Can you post and/or cite to some evidence for this claim? Because I'm a bit skeptical that there's much difference. High-quality, really comparable data is difficult to find, but there are some that I consider reasonably relevant to the issue of Dmax on matte papers from Keith Cooper, who I regard as quite knowledgeable and reliable.

Canon Pro-2000
Pinnacle Velvet 275gsm Dmax 1.663
https://www.northlight-images.co.uk/canon-imageprograf-pro-2000-printer-review/

Canon Pro-1000
Canon Fine Art Photo Rag Dmax 1.662
https://www.northlight-images.co.uk/canon-pro-1000-printer-review/

Epson P5300 / P5370
Hahnemuhle natural line Hemp paper 1.661
https://www.northlight-images.co.uk/epson-sc-p5300-printer-review/

Yes, I realize that those two Canons use the prior-generation inkset relative to the Pro-4600, but I haven't seen any claim, much less credible evidence, that Dmax has gotten worse. Yes, I realize the P5300 / P5370 uses a slightly different inkset from the P9750, but I haven't seen any claim, much less credible evidence, that Dmax varies much between them. Yes, those are three different papers, and Dmax probably varies a little from one to the other.

So I'd welcome better data, but I remain skeptical of the claim.

[JCD]

This is from the Epson Io Group. I tried to send the link but it requires the recipient being a member. This is from a friend Walker Blackwell , a highly regarded fine art printmaker for the last 23 years or so. He’s tech specialist and ink developer etc. He has the new Canon 60” and sent us his test results. I just copied and pasted part of it
In the end it gets into lots of dmax measurements of tons of papers. By the way he likes the 6600 but there are issues, as you can read below.


I will depart from Walker on two points. It seems to me the comments he’s making also about Epson printers are
not about the 9570 with it’s particular software and firmware and two level head design. He doesn’t have access to a 9570 or 7570.

There are no bronzing or gloss differential residual effects on Platine with the 9570.( there was with the P9000 and earlier.)   Platine is my favorite most used paper and neutral bw on these printers is nothing short of amazing with zero issues. I’ve also had perfect gloss rendition on the HP Z3200, but not my old Canon. The second is dither. I will believe it when I see it that Canon is capable of the 9570 dither at unidirectional! Again he’s not been testing the new LF Epsons, if even the P900.

Again, I’m not knocking the 4600. For that price and the  performance and real durability, full set of 350 ml inks  right now the 4600 is a steal,  as long as they don’t have issues down the line that we are unaware of. That would be rare for Canon. They have a history of no headaches.


——————————————————————
Quote:

Canon 6600 update. There’s good and there’s bad like every printer. I’m going to give you my non-gloss-optimized opinion.

Attached below is a raw un-smoothed LAB 128 patch test (MO measured) from abw mode in Standard contrast  neutral toning on Canson Platine (admittedly an old sheet I pulled out of a flat file). Platine has always been my problem-child for gloss optimizer. Photos of the 6600 bronzing in ABW mode below: not great Bob. I also included a screenshot of a target luminance contrast (green line low on left, high on right) and measured luminannce (blue line). Standard mode ABW on the Canon 6600 is very light and in-general Epson ABW and Canon ABW are way off as far as Luminance/contrast screen-to-print-match. That is easily overcome with a bit of math though . . .

At any rate, there is still a lot that canon needs to do to make the GCO work well on non-baryta gloss papers like Platine. We put a ton of energy into Piezography GCO to work better on these papers but I can’t imagine having to do this without the ability to add resin (because it’s a thermo head not a piezo head, it can’t print with that type of encapsulation). Seems like an incredibly hard engineering challenge to me.

That said, the baryta and poly gloss papers work exceptionally well and I have had really no issues with gloss bronzing on color prints. The over-all dither quality is BETTER than epson too now. I know that is crazy but I see no dots in highest quality unidirectional mode (it’s also a slower print than highest quality epson so that’s a caveat).

Over-all I think for straight monochrome, Piezography Pro is still king for gloss and matte fine art printing, but for a 60" monochrome gloss print (baryta paper) or 60” color gloss print I would go with this printer hands-down. Having the white of the paper = the gloss of the ink is critical to matching c-print or silver level uniformity in my opinion.

I guess I can start testing with Ergosoft at some point. This should allow me to increase the GCO behind its prescribed amount. In the case of Piezo, doing this helped eliminate that red bronzing from peaking out.

Another big thing here is the lights I use (GVM grids) will show bronzing way more than regular gallery lights or window light . . . so what you see below is the worst paper and worst light for bronzing.

///

On another note, I am severely disappointed with the matte black dmax. First print was L* 14.4 on a cheap paper. After running a lot of black it went all the way to L* 18. This is a critically f*cked thing and could potentially doom the entire canon market for this type of printer permanently if not addressed. Honestly, if they don’t address it with me and other adopters promptly, they could well deserve to exit the market with a little fart wiff and nobody would miss them. I will print out maybe 100mL of this expensive ink next week and see if it settles down into a somewhat reasonable density. If not I may have to use some other ink in the MK channel. But for real, in 2024 it’s like missing the ball and hitting the umpire smack in the face.



quote author=NAwlins_Contrarian link=topic=143780.msg1260253#msg1260253 date=1720290612]
Can you post and/or cite to some evidence for this claim? Because I'm a bit skeptical that there's much difference. High-quality, really comparable data is difficult to find, but there are some that I consider reasonably relevant to the issue of Dmax on matte papers from Keith Cooper, who I regard as quite knowledgeable and reliable.

Canon Pro-2000
Pinnacle Velvet 275gsm Dmax 1.663
https://www.northlight-images.co.uk/canon-imageprograf-pro-2000-printer-review/

Canon Pro-1000
Canon Fine Art Photo Rag Dmax 1.662
https://www.northlight-images.co.uk/canon-pro-1000-printer-review/

Epson P5300 / P5370
Hahnemuhle natural line Hemp paper 1.661
https://www.northlight-images.co.uk/epson-sc-p5300-printer-review/

Yes, I realize that those two Canons use the prior-generation inkset relative to the Pro-4600, but I haven't seen any claim, much less credible evidence, that Dmax has gotten worse. Yes, I realize the P5300 / P5370 uses a slightly different inkset from the P9750, but I haven't seen any claim, much less credible evidence, that Dmax varies much between them. Yes, those are three different papers, and Dmax probably varies a little from one to the other.

So I'd welcome better data, but I remain skeptical of the claim.
[/quote]

[Czornyj]

1) New MBK in PRO-4600 is improved and should reach L* 14-15 without a problem
2) Paper feed adjustment is critical to get optimal screening from Canon Pros, and apparently it wasn't done right in this case

[Asael]

My question about the 4600 is how lightfastness is it? "improved" and 200 years (is that dark storage??) Are not an answer, and the fact that this information is not clear concerns me.

Epson on the other hand, every time I buy one, I swear I will never buy another...

[Czornyj]

My question about the 4600 is how lightfastness is it? "improved" and 200 years (is that dark storage??) Are not an answer, and the fact that this information is not clear concerns me.

Epson on the other hand, every time I buy one, I swear I will never buy another...

It should be at least as good as newest Epson inkset (and even a bit better).

[MHMG]

It should be at least as good as newest Epson inkset (and even a bit better).

That would be nice, but Canon apparently has opted to base its LUCIA PRO II longevity claims on one paper (Platinum Pro) and with internal testing using the JEITA CP3901 standard, rather than relying on an independent testing laboratory. Worse yet, CP3901 is already a highly antiquated testing protocol even if it was published recently.

One has to pay JEITA to get the full published standard, but this overview  (https://home.jeita.or.jp/tsc/std-pdf/CP3901.pdf) tells me all I need to know.

To cut to the chase:  The JEITA test standard relies on obsolete densitometric testing criteria and with only CMYK colorants tested (no reds, greens, blues, etc., nor important color blends such as skin tones). CP3901 also specifies a longevity prediction based on 250 lux for 12 hours per day. 250 lux is essentially half of the assumed light intensity on display compared to Wilhelm's predictions, hence claimed "years on display" will double compared to Wilhelm predictions for no other reason than the change in this particular standard's assumed light level on display. The endpoint criteria are also different compared to those used by Wilhelm, e.g., more forgiving in key aspects, but this is largely academic because the test simply isn't good enough for multi-colorant ink sets no matter what endpoint criteria are chosen.

The LUCIA PRO II ink set may well have significant improvements over the earlier LUCIA formulations. Initial image quality and scratch resistance improvements should be easily visible to the enduser, but print longevity claims cannot be meaningfully confirmed or disputed by the consumer. Hence, if longevity claims are to be made, it's important for the industry to police itself a lot better than it seems to be doing these days!

Mark McCormick

[JCD]

I love these comments like “ It should be better” than Epson etc. They told me it should be much better, with no evidence to backup that statement.

If we’re lucky we’ll get back to where the IPF 8300 was 14 years ago, but with better gamut.

 If what Mark just said is accurate ( he of all people should know) a 250 lux exposure intensity method is like the old Kodak fake test and is pretty much meaningless in the real world. Just another way to throw around “ years of display “ figures. We’ll see .

I asked the Canon tech guru during the unveiling event webinar why they would wait so long after the printer has been put on the market to give real test data, especially since they are using this great improvement as one of the major marketing points of this series.

I said when can we expect published data on a variety of media to compare directly with Epson and HP. He wasn’t sure, and said “ these things take a long time.” Which is true, but not a logical excuse.  He spent half of the webinar talking about the environmentally friendly packaging. And said almost nothing about monochrome and little about what papers the revised gloss enhancer works on or the MK dmax.

John


quote author=MHMG link=topic=143780.msg1260268#msg1260268 date=1720470068]
That would be nice, but Canon apparently has opted to base its LUCIA PRO II longevity claims on one paper (Platinum Pro) and with internal testing using the JEITA CP3901 standard, rather than relying on an independent testing laboratory. Worse yet, CP3901 is already a highly antiquated testing protocol even if it was published recently.

One has to pay JEITA to get the full published standard, but this overview  (https://home.jeita.or.jp/tsc/std-pdf/CP3901.pdf) tells me all I need to know.

To cut to the chase:  The JEITA test standard relies on obsolete densitometric testing criteria and with only CMYK colorants tested (no reds, greens, blues, etc., nor important color blends such as skin tones). CP3901 also specifies a longevity prediction based on 250 lux for 12 hours per day. 250 lux is essentially half of the assumed light intensity on display compared to Wilhelm's predictions, hence claimed "years on display" will double compared to Wilhelm predictions for no other reason than the change in this particular standard's assumed light level on display. The endpoint criteria are also different compared to those used by Wilhelm, e.g., more forgiving in key aspects, but this is largely academic because the test simply isn't good enough for multi-colorant ink sets no matter what endpoint criteria are chosen.

The LUCIA PRO II ink set may well have significant improvements over the earlier LUCIA formulations. Initial image quality and scratch resistance improvements should be easily visible to the enduser, but print longevity claims cannot be meaningfully confirmed or disputed by the consumer. Hence, if longevity claims are to be made, it's important for the industry to police itself a lot better than it seems to be doing these days!

Mark McCormick
[/quote]

[Asael]

The lack of information and it's vagueness by Canon makes me not sure about the 4600. More so, when the previous generation did not have the lightfastness I needed, and that in the past Canon did provide good information. So IMO, their is a problem, either with their marketing or the lightfastness

[MHMG]

1) New MBK in PRO-4600 is improved and should reach L* 14-15 without a problem
2) Paper feed adjustment is critical to get optimal screening from Canon Pros, and apparently it wasn't done right in this case

Interesting. the Canon USA site indicates that 11 of the 12 inks are reformulated...the only one remaining the same is the MBK!
Here are some quotes from Canon's site https://www.usa.canon.com/shop/p/imageprograf-pro-4600?color=Black&type=New. First, note the bullet point "Reformulated LUCIA PRO II Ink ***" and also note the three asterisks. Second, note the Product disclaimer close to the bottom of the page "*** The Matte Black Ink is the same ink formulation as the imagePROGRAF GP-4000/2000".

So, all other things being equal, I would expect the new printer will show the greatest improvement on papers requiring the PBK photo black but not much if any on matte papers using MBK. However, other things may not be equal because the new driver may be departing from 100% GCR at DMax (RGB = 0,0,0 in the image file) by feathering in some other reformulated colorants or even some PBK when printing on matte papers. I have no insider knowledge, so I can only speculate, but it isn't out of the question that Canon engineers may have tweaked the Dmax levels on typical fine art matte media by some means other than reformulating the MBK ink.

cheers,
Mark

[Czornyj]

Interesting. the Canon USA site indicates that 11 of the 12 inks are reformulated...the only one remaining the same is the MBK!
Here are some quotes from Canon's site https://www.usa.canon.com/shop/p/imageprograf-pro-4600?color=Black&type=New. First, note the bullet point "Reformulated LUCIA PRO II Ink ***" and also note the three asterisks. Second, note the Product disclaimer close to the bottom of the page "*** The Matte Black Ink is the same ink formulation as the imagePROGRAF GP-4000/2000".

MBK was alredy reformulated in GP-4000/2000 (and PRO-300) as well

[Czornyj]

To cut to the chase:  The JEITA test standard relies on obsolete densitometric testing criteria and with only CMYK colorants tested (no reds, greens, blues, etc., nor important color blends such as skin tones). CP3901 also specifies a longevity prediction based on 250 lux for 12 hours per day. 250 lux is essentially half of the assumed light intensity on display compared to Wilhelm's predictions, hence claimed "years on display" will double compared to Wilhelm predictions for no other reason than the change in this particular standard's assumed light level on display. The endpoint criteria are also different compared to those used by Wilhelm, e.g., more forgiving in key aspects, but this is largely academic because the test simply isn't good enough for multi-colorant ink sets no matter what endpoint criteria are chosen.

Mark, I really don't think they used densitometric testing while even for profiling they use spectroradiometric measurements. I'm also quite confident that while internal testing they checked Epson inks by themselves, and didn't compare apples to oranges, or Wilhelm's test results - like all other companies Canon is not ideal and makes weird decisions and mistakes, but at the end of the day they tend to draw conclusions, fix imperfections and really don't seem to be idiots in color science area.

[MHMG]

Mark, I really don't think they used densitometric testing while even for profiling they use spectroradiometric measurements. I'm also quite confident that while internal testing they checked Epson inks by themselves, and didn't compare apples to oranges, or Wilhelm's test results - like all other companies Canon is not ideal and makes weird decisions and mistakes, but at the end of the day they tend to draw conclusions, fix imperfections and really don't seem to be idiots in color science area.

I can only go by what Canon's marketing folks have chosen to cite as evidence to support their print longevity claims. Canon cites JEITA CP3901 (not the WIR 3.0 method nor any colorimetric measurements) as its proof for improved print longevity of the new inks.

If Canon management had asked me (they didn't), I would have recommended that in the absence of a meaningful ISO light fastness testing standard for prints and photographs, Canon should simply publish a comparative head-to-head evaluation of LUCIA PRO versus LUCIA PRO II using delta E measurements taken from a low patch count color profiling target. And run a few representative third party papers, i.e., both matte and glossy papers from Canson, Hahnemuhle, etc.  Likewise, if Canon wants to claim comparable performance to Epson and/or HP, then include those sets in the publication as well. This approach would also exercise important color blends, and experienced printmakers would have some sense as to what the Delta E changes mean, but at the very least such an approach would offer valid proof that the new ink set outperforms the older one by some measure. As it stands with JEITA CP3901, the fade resistance of many of the 12 LUCIA PRO and PRO II inks doesn't even get factored into the test result. Canon R&D undoubtedly knows how they perform, but it's disingenuous for Canon marketing to cite a test standard that leaves out so much of the overall performance of the set. It's the same problem with the WIR 3.0 test method. Densitometry-based tests are incomplete for multi-colorant ink sets where complex fading interactions between inks can sometimes happen. BTW, spectral measurements can easily be converted to densitometric values (not the other way around), but at the end of the day and no matter how extensively the R&D team's chemical and engineering efforts were to improve performance of the new ink set, the Canon marketing team has once again dropped the ball, IMO.

All that said, the colorimetry-based I* metric which was published in late 2004 is superior to delta E for color appearance evaluations concerning prints and photographs because it measures the critically important property of tonal contrast whereas delta E does not. It also properly weights hue and chroma changes for colors seen in the context of real images whereas even DE2000 does not. Also, with over 300 ink and media tests I conducted between 2007 and 2022 and all available for free in the Aardenburg light fade test results database, I demonstrated its superiority to both densitometry and the various flavors of Delta E.  When I finally shut down the light fade testing program at Aardenburg Imaging & Archives in 2022, I offered the entire Aardenburg light fade testing protocol to the ISO committee working on print permanence standards at no charge. I would have volunteered to help the ISO committee members turn the Aardenburg testing protocol into an ISO test standard. The committee said thanks, it would get back to me.

I never heard from the ISO committee after that. It is what it is.

The print permanence issues of concern to me now are the light-induced low-intensity yellowing/staining (LILIS) problems affecting all papers containing TiO2 and OBAs, and the phenolic yellowing problem which affects almost all fine art media nowadays due to the alkalinity and porosity of the microporous coatings. My research into both problems has come far enough along that I now believe the media manufacturers are in a position to offer new media in their product lines which greatly reduce these yellowing issues if they care to.  However, both issues fly mostly under the radar in the printmaking community, so the media manufacturers aren't feeling any customer pressure to improve products they already market to us as "archival".

cheers,
Mark