p.2 #1 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
@BeatX, if there's good detail in the image file but the print is not reflecting the same quality, it would suggest a problem with the printing. Have you ever printed anything of similar size using this same lab? Many "labs" have 60 inch printers. Relatively few do exceptionally good printing.
p.2 #2 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
There a lot of contradictory opinion and thoughts here. First cheating physics, of course you can't but despite your words, that seems to be exactly what you are trying to do. Mathematically, using a 40mp sensor to capture more data than a 26mp, is not cheating, that is physics. Heavy sharpening and upscaling are attempts to cheat physics. I have not jumped into the AI/upscale/fix it software much. I get that it could be useful to address high ISO noise or even some amount of sharpening but you are using software algorithms to fill in more data that wasn't in the original, I think you have missed their best application. To take a well exposed and sharp photo and process it to get "more" of something is not what I see the tools ideal for. I would stick to using them for fixing issues not sharpening to hyper-reality, and risking over-processed funky looking images.
Your perspective on sensors is a little inconsistent. You basically say that an XT cropped down to 26mp is going to have the same detail printed at the same size as a 40mp one. Not in my experience. Sure you can crop, but the image always degrades a little. You also seems to suggest that upscaling a 26mp image could improve it, yet you just said it past the point of diminishing returns. If the sensor can't do it, software interpolation isn't going to help.
I don't have a 26mp currently but I do have the XT5 and Gfx50S. The Gfx is in a different league but I don't think it has anything to do with more resolution. In fact I see very little difference in resolved detail between the two. More about realistic rendering and gorgeous color to my eyes. Very little post-processing needed but if you do, there's crap loads of dynamic range. The Gfx can handle higher ISO's as well.
Printing - I have printed several XT files with good success at 20"x30" (50x76cm) and didn't feel that the limit was that close, could go bigger. Simple processing in Capture One, to adjust highlights, shadows, exposure but aiming to match the scene that I could see that day. Also, the media printed on matters quite a bit as well. The 20"x30" photo paper prints have better detail than another print onto 16"x20" canvas. The canvas is beautiful, has a great feel to it. But it doesn't do fine detail all that well. Also, when I have surveyed the different focal lengths used on my favorite prints and do not find that the highest resolving lenses stole the show. But then to me, subject matter trumps pixel peeping detail, and most of my favorites shots were done on the fly with a hand held zooms that YouTubers want to hate - 10-24, 16-80. I'm not dissing primes, there's just a time and a place. Regarding that Viltrox 13mm that is a YT reviewer favorite, I think it is the most over-rated lens I have tried. Its great for the price but I would rather shoot my 10-24 at 13mm, than the Viltrox. I only keep it around for the speed, keep saying I'm going to try it at astro before I sell it. YT reviews are nice, but I find they are often more interested in stoking a controversial idea more than useful info. I guess saying its a mediocre lens but a great price isn't going to pull in subscribers.
p.2 #4 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
gdanmitchell wrote:
Indeed, you cannot personally evaluate the quality of that print in the form I presented it here. It is difficult, I hear, to upload 40” prints to FM.
It isn’t that I’m “happy with this print.” It was a test to see what a print at that size would look like. It looks very good.
Exactly. This poster tells us that a 40MP x-trans sensor camera is limited to print of about the 12” x 18” size. That’s simply ridiculous.
Not exactly limited, but I mean that there is hardly no difference in fine details in prints A2 size between 26Mpix and 40Mpix sensor.
Again, lenstip.com resolution measurement results shows, that there is 10% real gain of sharpness in MTF function between X-T4 and X-H2 sensors https://www.optyczne.pl/482.4-Test_aparatu-Fujifilm_X-H2_Rozdzielczo%C5%9B%C4%87.html
So I run simple experiment.
I downscaled my original image to 26Mpix in C1, then I upscaled it again to original 40Mpix.
Here are results:
Original file:
downscaled to 26Mpix and upscaled to 40Mpix:
As You can clearly see, even such destructive processing basically don't change amount of fine details.
I mean, there are some very (and I mean very) little differences, which - again - would confirm lenstip.com X-Trans V 40Mpix sensor measurement (10% real gain of sharpness in MTF function between X-T4 and X-H2 sensors)
So it seems like, printing from X-Trans V 40Mpix sensor, is more-less like printing from former 26Mpix sensor (in practice it works somewhat like software upscaling)
My conclusion: A2 size prints from Fujifilm X-Trans V 40Mpix sensor is beyond its ability to resolve fine details for such size of print paper (even if number of megapixels might seem to be enough)
Because in reality, like in many shown examples in this thread - X-Trans V 40Mpix sensor has very symbolic gain when it comes to increasing resolution (amount of fine details) in real world usage.
Please, someone correct me if I'm wrong with some good photo examples
That assumes sensor can achieve 100% of the theoretical limit and it won't. However, a lens that can do 94 lp/mm is an excellent lens but with today's technology pretty achievable. A lens that can do 164 lp/mm is a completely different story. That is a world class lens and the cost to make it is going to be high and the design complex.
Let's be generous and assume the sensor can achieve the theoretical. You still have mm, which is a size measurement.
Fuji GFX 50 = 1441.02mm2 (43.80mm x 32.90mm)
Fuji X-T5 = 366.6mm2 (23.50mm x 15.60mm)
43.80 x 94 = 4,117.2
32.9 x 94 = 3,092.6
23.50 x 164 = 3,854
15.6 x 164 = 2,558.4
However, I wouldn't get caught up in all that because generally almost nobody prints at sizes where that is going to make a difference. What you have is capable of much larger prints. Are you sure you are using a quality print shop? The other simpler explanation is you didn't see a difference between 24 or 40 MP APS-C is because you printed too small at 16x23 inches (40x60cm)....Show more →
cbass thank You for very informative YT videos (after watching them I was rally surprised of conclusions)
Food for thoughts, heh..
Also, thx for very informative post with detailed calculations, but unfortunelly it's beyon my understanding (those calculations and its meaning)
Although, what I do undrestand is fact, that when it comes to resolving fine details, most limiting factor is lens ability to resolve more resolution?
p.2 #6 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
jeffbuzz wrote:
@BeatX@, if there's good detail in the image file but the print is not reflecting the same quality, it would suggest a problem with the printing. Have you ever printed anything of similar size using this same lab? Many "labs" have 60 inch printers. Relatively few do exceptionally good printing.
No, I've never printed in A2 size from this lab, only A3
But my experiment with downscaling file is worry me. It shows clearly that there is no real difference when it comes to fine details between 26 and 40Mpix (ergo, printing in A2 from X-Trans V sensor is like printing from former 26Mpix)
p.2 #7 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
BeatX wrote:
No, I've never printed in A2 size from this lab, only A3
But my experiment with downscaling file is worry me. It shows clearly that there is no real difference when it comes to fine details between 26 and 40Mpix (ergo, printing in A2 from X-Trans V sensor is like printing from former 26Mpix)
Are you sure the lab is not downsizing your file? Many labs have resolution limits per print size. They might only print 300dpi up to A4. A3 might be scaled to 250dpi, A2 @ 200dpi and so on. Are you 100% sure of what this lab is doing? Some labs simply cannot print A2 @ 300dpi because they don't have a good enough computer or software. There are legitimate reasons to print A2 @ lower resolution. Mainly because the expected viewing distance would be measured in meters, not centimeters. You may need to explicitly request higher res printing and likely pay more to get it.
p.2 #8 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
jeffbuzz wrote:
Are you sure the lab is not downsizing your file? Many labs have resolution limits per print size. They might only print 300dpi up to A4. A3 might be scaled to 250dpi, A2 @ 200dpi and so on. Are you 100% sure of what this lab is doing? Some labs simply cannot print A2 @ 300dpi because they don't have a good enough computer or software. There are legitimate reasons to print A2 @ lower resolution. Mainly because the expected viewing distance would be measured in meters, not centimeters. You may need to explicitly request higher res printing and likely pay more to get it....Show more →
I was wondering something similar. I don't print, so you shouldn't trust anything I say. I have labs print my stuff, so I have no experience here. However, 300 dpi seems to be the resolution limit of most printers. If that is correct, then details are going to be limited to 300 Dots Per Inch at best. That means the only way you are going to see more details is by increasing the print size.
This is a simplification because not only are there different printers, but also different print processes. RGB vs CMYK. I think the photo paper that is selected affects the color gamut and maybe even the resolution supported. I am not confident in that last statement.
p.2 #9 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
BeatX wrote:
cbass thank You for very informative YT videos (after watching them I was rally surprised of conclusions)
Food for thoughts, heh..
Also, thx for very informative post with detailed calculations, but unfortunelly it's beyon my understanding (those calculations and its meaning)
Although, what I do undrestand is fact, that when it comes to resolving fine details, most limiting factor is lens ability to resolve more resolution?
My understanding is that it's not that simple. You can't just say a lens is limited in resolution. If you look at lens rentals, they say it's a combination of sensor and lens. Let's say at base ISO a sensor can achieve 90% contrast. Let's say a lens at 60 lp/mm can resolve 60% contrast. The total contrast will be .9 x .6 = 0.54 with a theoretical maximum of 1. Take the same lens and put it on a sensor that can do 98% you get .98 x .6 = 0.588, which is an improvement. Buy a better lens that can do 80% and you get .9 x .8 = .72, which is also an improvement.
How lenses start to behave at something like 200 lp/mm over 40 or 60 lp/mm is unpredictable:
A relevant snippet from that article so you aren't stuck reading it all:
One of the reasons I was excited about testing new lenses was that the Sony 135mm f/1.8 GM had the best normal-range MTF scores we’d ever seen, besting the Sigma 135mm f/1.8 Art. It seemed logical that it might also best the Sigma in ultra-high resolution tests. Reality 1, Logic 0. We tried the lens at f/4, but it was actually a bit better at f/5, which we show below. At high frequency, it’s not as good as the Sigma.
The bottom line is that a bigger sensor has an advantage. It's not just the number of pixels. The resolution has been so good the last few decades that for 95% of people you shouldn't need more megapixels to print large. Unless you are running a gallery you aren't going to have the wall space in your house. All those extra details you pixel peeped on your monitor are going to be lost on anything but a massively large print. Save your money. Youtubers make money from affiliate links when you buy stuff. They get to go on special trips paid for by manufacturers for promoting the newest gear.
You don't need to understand all the technical stuff. There is a concept of diminishing returns. That means that when the reviews say it's a minor step forward that is not surprising.
It looks like at 300 DPI the dot pitch is 85µm. That's much larger than what a sensor can record. Even 4000 DPI is larger than the pixel pitch of these modern sensors.
p.2 #11 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
SGinNorcal wrote:
There a lot of contradictory opinion and thoughts here. First cheating physics, of course you can't but despite your words, that seems to be exactly what you are trying to do. Mathematically, using a 40mp sensor to capture more data than a 26mp, is not cheating, that is physics. Heavy sharpening and upscaling are attempts to cheat physics. I have not jumped into the AI/upscale/fix it software much. I get that it could be useful to address high ISO noise or even some amount of sharpening but you are using software algorithms to fill in more data that wasn't in the original, I think you have missed their best application. To take a well exposed and sharp photo and process it to get "more" of something is not what I see the tools ideal for. I would stick to using them for fixing issues not sharpening to hyper-reality, and risking over-processed funky looking images.
Your perspective on sensors is a little inconsistent. You basically say that an XT cropped down to 26mp is going to have the same detail printed at the same size as a 40mp one. Not in my experience. Sure you can crop, but the image always degrades a little. You also seems to suggest that upscaling a 26mp image could improve it, yet you just said it past the point of diminishing returns. If the sensor can't do it, software interpolation isn't going to help.
I don't have a 26mp currently but I do have the XT5 and Gfx50S. The Gfx is in a different league but I don't think it has anything to do with more resolution. In fact I see very little difference in resolved detail between the two. More about realistic rendering and gorgeous color to my eyes. Very little post-processing needed but if you do, there's crap loads of dynamic range. The Gfx can handle higher ISO's as well.
Printing - I have printed several XT files with good success at 20"x30" (50x76cm) and didn't feel that the limit was that close, could go bigger. Simple processing in Capture One, to adjust highlights, shadows, exposure but aiming to match the scene that I could see that day. Also, the media printed on matters quite a bit as well. The 20"x30" photo paper prints have better detail than another print onto 16"x20" canvas. The canvas is beautiful, has a great feel to it. But it doesn't do fine detail all that well. Also, when I have surveyed the different focal lengths used on my favorite prints and do not find that the highest resolving lenses stole the show. But then to me, subject matter trumps pixel peeping detail, and most of my favorites shots were done on the fly with a hand held zooms that YouTubers want to hate - 10-24, 16-80. I'm not dissing primes, there's just a time and a place. Regarding that Viltrox 13mm that is a YT reviewer favorite, I think it is the most over-rated lens I have tried. Its great for the price but I would rather shoot my 10-24 at 13mm, than the Viltrox. I only keep it around for the speed, keep saying I'm going to try it at astro before I sell it. YT reviews are nice, but I find they are often more interested in stoking a controversial idea more than useful info. I guess saying its a mediocre lens but a great price isn't going to pull in subscribers....Show more →
It looks like at 300 DPI the dot pitch is 85µm. That's much larger than what a sensor can record. Even 4000 DPI is larger than the pixel pitch of these modern sensors.
So, You mean there is no printer or technology to print on photo paper, to resolve all this fine details and resolution than lets say.. GFX platform can resolve?
Only solution is to publish images on big screens with ultra high resolution? (like 8k)
p.2 #13 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
jeffbuzz wrote:
Are you sure the lab is not downsizing your file? Many labs have resolution limits per print size. They might only print 300dpi up to A4. A3 might be scaled to 250dpi, A2 @ 200dpi and so on. Are you 100% sure of what this lab is doing? Some labs simply cannot print A2 @ 300dpi because they don't have a good enough computer or software. There are legitimate reasons to print A2 @ lower resolution. Mainly because the expected viewing distance would be measured in meters, not centimeters. You may need to explicitly request higher res printing and likely pay more to get it....Show more →
Hi jeff
I'm not sure if my lab is not downsizing files. I wasn't there, while my files has been pushed to printer.
Hard to say to be honest, nevertheless I printed a lot from this lab from my former X-S10, just in A3 size, and result was always excellent, rally nothing to complain about.
Tack sharp prints, corner to cornet with full details, beautiful colors and amount of contrast that I like.
But this is online photo lab, which was recommended to me by my mentor
Crazy passionate photographer with decades of experience (pixel peeper like maniac)
Half of his YT videos is only about rendering character from specific models of Voigtländer lenses lol
He do photos for living and he is living for photography.
He is GFX user and biggest fanboy of medium format I've ever seen.
BTW, this is his YT channel, but in polish language: https://www.youtube.com/@MaciejTaichman
Anyway, he is using this lab which he recommands me, where he do regular orders for big prints from his GFX 100S and I'm pretty sure, that if this lab meets his standards, that means it meats 99,9% photographers standards I've zero doubt in it.
p.2 #14 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
BeatX wrote:
So, You mean there is no printer or technology to print on photo paper, to resolve all this fine details and resolution than lets say.. GFX platform can resolve?
Only solution is to publish images on big screens with ultra high resolution? (like 8k)
I don't know. I don't do my own printing so my knowledge here is very limited. I found that information interesting and I put it out there hoping that someone with more knowledge and experience would comment on it.
p.2 #15 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
BeatX wrote:
No, I've never printed in A2 size from this lab, only A3
But my experiment with downscaling file is worry me. It shows clearly that there is no real difference when it comes to fine details between 26 and 40Mpix (ergo, printing in A2 from X-Trans V sensor is like printing from former 26Mpix)
I think there is a difference between the two files if you look closely. For example, the names on the boats in the centre of the frame (Volla II?) is more readable in the first file than the second. You also have to bear in mind that most screens display only a few megapixels. I think a 4k screen displays about 8 megapixels, so they are not the best way to test. Also if you are shooting hand-held and at higher ISO you will be losing fine detail anyway and atmospheric conditions can also play a factor in how much fine detail is displayed.
A better test will be something shot indoors at relatively close range with a lot of fine detail. I expect the 40mp sensor will resolve a bit more. I don't think it will be dramatic but I expect it will be noticeable at large print sizes of about 1m plus, if looking closely.
p.2 #16 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
Today I talked briefly to someone that has a printer. He is not a professional, but an amateur that prints. In order for a X-H2 file printed at 300 DPI to use every pixel (7728 x 5152) it would need to be slightly larger than 25 x 17 inches (65 x 43 cm). If you print larger than that, then the spacing of the dot would be greater and there would not be 300 Dots Per Inch. Assuming you don't upscale. Apparently, some labs or printer software will upscale for you while others will not complicating matters. What is not clear to me is what happens when you print smaller than 25 x 17. I assume that some of the information has to be thrown away, but I am interested in hearing an answer.
Then I decided that I would ask the lab that I use about the DPI they use and how they scale and then I get this response:
For our Photo Prints, we do not use inkjet printers. Instead, we produce Chromogenic Prints, also known as Silver Halide Prints. Our Photo Prints are made using a digital image and developed using a chemical process.
Silver Halide Photographic Prints are composed of three layers of gelatin, each containing an emulsion of silver halide (used as a light-sensitive material). The photo paper is exposed to light, and the image is infused into the paper through a chemical process. The resulting Prints have more depth, more vibrant colors, more details in highlights and shadows, are sharper, and are less susceptible to fading/damage as seen with Inkjet Photographic Prints. Inkjet Photographic Prints involve applying ink to just the surface of the paper!...Show more →
Well F&$K. It looks like the process they use is more in line with how film was developed, and DPI doesn't really apply there.
I have just been reminded why I decided to pay someone else to print for me. My brain has reached its maximum on this one for now.
p.2 #17 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
cbass wrote:
Today I talked briefly to someone that has a printer. He is not a professional, but an amateur that prints. In order for a X-H2 file printed at 300 DPI to use every pixel (7728 x 5152) it would need to be slightly larger than 25 x 17 inches (65 x 43 cm). If you print larger than that, then the spacing of the dot would be greater and there would not be 300 Dots Per Inch. Assuming you don't upscale. Apparently, some labs or printer software will upscale for you while others will not complicating matters. What is not clear to me is what happens when you print smaller than 25 x 17. I assume that some of the information has to be thrown away, but I am interested in hearing an answer.
Then I decided that I would ask the lab that I use about the DPI they use and how they scale and then I get this response:
Well F&$K. It looks like the process they use is more in line with how film was developed, and DPI doesn't really apply there.
I have just been reminded why I decided to pay someone else to print for me. My brain has reached its maximum on this one for now. ...Show more →
Some labs print that way, but most don’t. A few will offer a choice. Most are going to use inkjet printers.
The statement about print size that “uses every pixel” omits a critical data point: what ppi resolution is that?
At 300 ppi you’ll get a larger non-interpolated print than you would get at 360 ppi with the exact same image.
Are you sure you understand the difference between dpi (or dots per inch, which refers to how the printer lays down individual “drops” of ink) and ppi (or pixels per inch, which refers to the number of image pixels used for each linear inch of the print)? For example, I typically print at 360 ppi resolution on a printer that uses 2880 dpi.
I”m not certain what process your print shop is using but ppi (image resolution) is still relevant, as a digital light source is being used to project pixels onto the paper before development.
How much ppi resolution is enough? 300 ppi is pretty typical, though there’s an argument that 360 might be more appropriate for Epson printers, since their dpi settings are multiples of 360. But lower ppi values are also fine in many cases, especially in very big prints. A general rule of thumb is that if you interpolate to make a larger print you can go as low as around 200 ppi resolution and not see pixelation.
Contrary to the expectations of many, higher ppi values are arguably more necessary with small prints than with large prints. A large print is typically viewed from a somewhat greater distance than a small print, which even be held in the hand. Up close, in a small print, you might want higher resolution settings in order to accurately render small details at teh small print size.
p.2 #18 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
I think all here would benefit from buying and reading Jeff Schewe’s book, “The Digital Print.” There’s so much apparent confusion here as regards digital prints and printers, ppi, dpi, any given printer’s native resolution and how that interacts with what you “send” to that printer in ppi when printing - that it would be impossible in a post or two to tackle what seems to me to be some significant gaps in understanding how it all works.
Just a couple of points for the record. Modern digital printers are not RGB devices, they’re CMYK devices and there’s a conversion process that happens going from an RGB color space when sending data to the printer. Each printer brand/model has a “native resolution” for it’s print head, typically either 300 or 360. It’s best practice to send an image to the printer at native print head resolution. (On some printers, when the native resolution of the image at any given size would be much larger than 300 or 360, due to large file size, there can be improvement in the print quality by sending 600 or 720 to the printer with concomitant driver settings.) What the printer lays down in dpi isn’t as simple as it appears from the discussion in this thread. Variable size “dots” (droplets in tiny amounts at the picoliter scale) combine to make colors and lay down the image on paper. The sharpness of the print on paper depends on not only on the data “in the image” but also the type of paper, its receiver coating formulation and the amount of “spread” of the ink when it hits he paper (dot gain). Some print utilities can do upsampling or down sampling “on the fly” to send the printer an image at native print head resolution, and do a darn fine job of it.
p.2 #19 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
Rand47 wrote:
I think all here would benefit from buying and reading Jeff Schewe’s book, “The Digital Print.” There’s so much apparent confusion here as regards digital prints and printers, ppi, dpi, any given printer’s native resolution and how that interacts with what you “send” to that printer in ppi when printing - that it would be impossible in a post or two to tackle what seems to me to be some significant gaps in understanding how it all works.
Just a couple of points for the record. Modern digital printers are not RGB devices, they’re CMYK devices and there’s a conversion process that happens going from an RGB color space when sending data to the printer. Each printer brand/model has a “native resolution” for it’s print head, typically either 300 or 360. It’s best practice to send an image to the printer at native print head resolution. (On some printers, when the native resolution of the image at any given size would be much larger than 300 or 360, due to large file size, there can be improvement in the print quality by sending 600 or 720 to the printer with concomitant driver settings.) What the printer lays down in dpi isn’t as simple as it appears from the discussion in this thread. Variable size “dots” (in tiny amounts at the picoliter scale) combine to make colors and lay down the image on paper. The sharpness of the print on paper depends on not only on the data “in the image” but also the type of paper, its receiver coating formulation and the amount of “spread” of the ink when it hits he paper (dot gain). Some print utilities can do upsampling or down sampling “on the fly” to send the printer an image at native print head resolution, and do a darn fine job of it.
p.2 #20 · My opinion and thoughts about Fujifilm X-Trans V 40Mpix sensor
Great advice. Years ago I got the Luminous Landscape “capture to print” (or similar title) series of videos, and I learned a ton about all of this that continues to serve me well.
If one is serious about printing — even if sending stuff out to a service, but especially if you do your own printing — eventually you’ll really want to dig into this stuff and understand it more fully.
Admittedly, that isn’t for everyone. Many folks will legitimately prefer to _not_ go that deeply and to just send stuff out to a service. If one does that, it is still worth having a conversation with the service about the best way to format files for them. (It is also a great idea to at least get a small printer and work out what you want the file to look like before sending it off to the printer for a larger version.)
The print head “native resolution” thing is a bit more complicated that it being either 300 or 360. More accurately, the highest dpi resolution (based on the spacing of the ink droplets on the paper) is a multiple of one of those numbers. (I earlier mentioned Epson, since their printers — at least those I use and have used — have dpi resolutions that are multiples of 360, such as 720, 1440, 2880, and so forth. (I’ve often wondered if the old advice that you could go as low as 180 ppi with big prints is related to this.) So what you actually strive for is to send an image to the printer with a ppi resolution that is a whole number division of the printer’s dpi resolution.
But speaking of Jeff Schewe, if I recall correctly from that old video series I mentioned, he did some experiments and determined that because printer interpolation routines are actually quite good that you could just send almost any high resolution setting to the printer and it would look almost indistinguishable from what you’d get if you did your own pre-interpolation on your computer!
Your point about differences based on types of paper is also quite correct, though perhaps beyond the level of concern for all but the most serious printers. (Your point about dot gain almost sounds like it could be quoted from some stuff I’ve written!) Schewe recommended, and I’ve always done this, slightly “over-sharpening” the file at the final stage before sending it to the printer as a way to compensate for dot gain. If you want to get really picky — and I do — you will use different degrees of over-sharpening for different papers since their dot gain varies.
OK, we’re pretty far into the weeds here. Interesting stuff for some of us, but probably beyond what most people sending a few files out to a printer need to know.
Rand47 wrote:
I think all here would benefit from buying and reading Jeff Schewe’s book, “The Digital Print.” There’s so much apparent confusion here as regards digital prints and printers, ppi, dpi, any given printer’s native resolution and how that interacts with what you “send” to that printer in ppi when printing - that it would be impossible in a post or two to tackle what seems to me to be some significant gaps in understanding how it all works.
Just a couple of points for the record. Modern digital printers are not RGB devices, they’re CMYK devices and there’s a conversion process that happens going from an RGB color space when sending data to the printer. Each printer brand/model has a “native resolution” for it’s print head, typically either 300 or 360. It’s best practice to send an image to the printer at native print head resolution. (On some printers, when the native resolution of the image at any given size would be much larger than 300 or 360, due to large file size, there can be improvement in the print quality by sending 600 or 720 to the printer with concomitant driver settings.) What the printer lays down in dpi isn’t as simple as it appears from the discussion in this thread. Variable size “dots” (in tiny amounts at the picoliter scale) combine to make colors and lay down the image on paper. The sharpness of the print on paper depends on not only on the data “in the image” but also the type of paper, its receiver coating formulation and the amount of “spread” of the ink when it hits he paper (dot gain). Some print utilities can do upsampling or down sampling “on the fly” to send the printer an image at native print head resolution, and do a darn fine job of it.
