• Please use real names.

    Greetings to all who have registered to OPF and those guests taking a look around. Please use real names. Registrations with fictitious names will not be processed. REAL NAMES ONLY will be processed

    Firstname Lastname

    Register

    We are a courteous and supportive community. No need to hide behind an alia. If you have a genuine need for privacy/secrecy then let me know!
  • Welcome to the new site. Here's a thread about the update where you can post your feedback, ask questions or spot those nasty bugs!

Another newbie question

Margo Stewart

New member
Ok, here's something else I'm curious about. I generally edit all of my photos in DPP and/or PS. Someone told me that when I save it in PS I first need to go to "mode" and then select "8 bits/channel". I do know that if I don't select that then I don't have the option to save it as a jpeg. I don't know WHY I do that though. Other than the whole jpeg thing.

I also just noticed today that when do a "save batch" in DPP, it automatically pulls up 72 dpi. I read yesterday that for prints it should be 300 dpi. Is that right? And how do I know what the dpi is when I save it after editing in PS?

Lastly, I usually shoot in the biggest size I can so I can crop it. After a family vacation in May I edited all our photos and then had them printed. I was bummed to have several that I had cropped come back differently then how I had cropped them. For instance, I took this photo of my son.

May18tResized.jpg


And then I cropped and edited it. (You can see that I had focus issues here as well. As I mentioned in my other post...if I had the aperture set a little more open I think I would have had more in focus and not just the top of his head. ????)
May18uResized.jpg


But when I got the print back part of it was cut off. Something about the crop ratio or print ratio? Obviously I don't know. So where can I learn how small I can crop something and still have it print correctly?

Thanks!
Margo
 

Diane Fields

New member
Margo, I know you're probably pretty busy with a full family life, but if you really are serious about learning how to deal with PHotoshop, then I'd suggest a good book or 2, maybe some online learning. There are multitudes of tutorials about Photoshop.

One of my favorite books is always Martin EVening's PHotoshop for PHotographers--this may be more than you want to start with. There are lots of others--I'll have to think about this and maybe others have some of their favorites they would like to recommend. Also--if you are willing to spend the money and do online learning, lynda.com is a great place for good basic Photoshop learning. Here is a link to their free tutorials http://movielibrary.lynda.com/html/free.asp Another good book for your needs might be Deke McClelland's Adobe PHotoshop CS3, One on One---he also does a number of tutorials for lynda.com Some people like Scott Kelby's books too.

The photography itself is a different thing--more to learn. YOu seem to understand that there are such things as aperture, shutter speed, ISO, but need help in understanding how to use them (I noted that you got some good replies in the other thread).--so, maybe a book on exposure would be the best thing here. For starters, maybe the one I can think of off the top of my head. Its a large format paperback, I'm sure available in Barnes and Noble or Borders and certainly at Amazon.
"Understanding Exposure' by Bryan Peterson. If you are buying a used copy at Amazon, be sure its the revised edition.

The question about the cropping, resizing, print ratio will be based on what size prints you requested. Probably too long to go into here right now with the other questions, but that would be answered in the books mentioned.

Diane
 

Margo Stewart

New member
Thank you!

Diane, thank you so much for all the great resources you sent my way. This is the 2nd mention of "Understanding Photography" that I've heard of so I'm going to look into whether or not my local book store may have that available.

I have also found that a local technology school here in my town offers online courses. I'm wondering if that's the route I should take. I am obviously having trouble comprehending my camera! LOL

Margo
 

Diane Fields

New member
Diane, thank you so much for all the great resources you sent my way. This is the 2nd mention of "Understanding Photography" that I've heard of so I'm going to look into whether or not my local book store may have that available.

I have also found that a local technology school here in my town offers online courses. I'm wondering if that's the route I should take. I am obviously having trouble comprehending my camera! LOL

Margo

I know our local community college offered Photoshop courses online and people could learn at their own pace. That would be a good way to go to get started probably--and a lot more convenient than actually physically going to the school for the classes. Online learning is really making so much available for those that could never 'connect' for the actual classes.

Diane
 
Margo,
jpg format is 8 bit per pixel.

I also just noticed today that when do a "save batch" in DPP, it automatically pulls up 72 dpi. I read yesterday that for prints it should be 300 dpi

DPI (dot Per Inch) is an output resolution. It make sense only whan you lay down pixels, for example printing.

Images are grid of numbers, for red green and blue channel. Every point in the grid is a pixel. A pixel does not have physical size until you render it.

Image have a width and height measured as number of pixels, so no physical dimensions.

Now 72 dpi question.
In many image file format (jpg,tif,bmp,..) you can set an information of PPI/DPI (PPI is the correct term).
This info is for documentation purpose, it doesn' make any sense:
the image is wwww pixel in width and hhhh pixels in height..... not more not less.

The value stored in the file is ignored by every application.

To send 300 PPI file you have to build it depending on paper dimensions.

Forget the 300 PPI urban legend.

Sorry for the fast reply, but i'm an in hurry.
If my words appear confused, don't hesitate to ask.


Giovanni
 

Margo Stewart

New member
By Giovanni, I think I got it! ;)

To send 300 PPI file you have to build it depending on paper dimensions.

Thanks Giovanni! Actually, it seems to make sense the way you put it. And from what I think you said, it doesn't necessarily need to be 300PPI.

The only thing that confused me is the quote above. But I gather I'll perhaps learn it when my new Photoshop book gets here. ???

Margo
 

Anil Mungal

New member
But when I got the print back part of it was cut off. Something about the crop ratio or print ratio? Obviously I don't know. So where can I learn how small I can crop something and still have it print correctly?

Hi Margo,

The top image is 600x400 pixels. As a ratio it can be represented by 600:400, which reduces to a 3:2 aspect ratio, which evenly fits in a 6x4 print (or a 3x2 print, or a 9x6 print, etc). The bottom images is 600x507 pixels, which is approximately 6:5 as a ratio. What does that mean? .... It would make a great 6x5 inch print .... but to make a 6x4, you'd have to crop off a bit of height.

In Photoshop, use the Crop Tool (C) on your top image and fill in 6 inches and 4 inches and 300 pixels per inch in the settings and crop away. This will force you to crop in the aspect ratio you want, and you'll be in control of the crop (rather than your lab).
 

Margo Stewart

New member
Yes!

No THAT makes sense! I think I hear all these numbers and just figure it's too much like math. Somehow I got kids that are good at math. Not sure what parent they got THAT from. Definitely not me! LOL

Anyway, thanks for the info. It honestly does make more sense now. Sort of one of those "duh!" minutes!

Now just tell me you'll all remember my lame-o questions this time next year when I have (hopefully) dramatically improved! LOL

Margo
 

Todd harrison

New member
interesting! Yeah I printed out 72 dpi photos and they have turned out great. but it makes sense about what you guys are say. but to me I always says it's better to be big than to be too small. at least if you have a bigger image when you take the picture you can make it any size you want. but a smaller size is great for email and other stuff. I'll have to get used to PPI I'm used to say dpi. interesting I never heard it like that before.
 
300dpi is probably from the fact that most digital minilabs print at a resolution in the region of 300dpi. If you choose to have photographs printed by a photolab, the computer(s) contolling the minilab will automatically resize the image to an appropiate size for the print, obviously the greater number of pixels the computer starts out with, the better the final print will be (up to a point though: for a 4x6 inch print at 300 dpi no more than 1200x1800 pixels (~2.2MP) will be required).

As for cropping, setting the print dimensions in the options bar along the top when the crop tool is selected will help to achieve the desired image in the final print. If you are sending the photographs to a photolab to be printed, I suggest not to crop less than 300dpi because the image will be interpolated if it does not have sufficient pixels to print from - this may result in visible deteration in the quality of the image (in the form of pixelation).

This tutorial should help with sizing with the crop tool:
http://artist.tizag.com/photoshopTutorial/croptool.php
 

Dierk Haasis

pro member
The Printing Confusion

300 dpi is the same as 150 lpi [lines per inch], which in turn is the typical high quality relative resolution for magazines. 72 ppi [pixels per inch] is a remnant of older monitors, particularly Apple, run at that relative resolution; Windows native RR was 96 ppi, though it could always be set higher or lower.PPI and DPI are almost always used interchangeably. As is already posted, the RR put into the image metadata is only meaningful if you print without setting a specific output resolution, and only if the program you print from reads it.

What you want for printing is a good trade-off between [necessary] quality and sensible data volume. 300 dpi is all too often overkill. Unfortunately printing is still something like Black Magic, with only very few people able to delineate the issues in a way the regular gal* understands it. One of them is Mike Chaney of QImage fame [writes a monthly column at Steve's Digicams ] and Harald Johnson, whose book Mastering Digital Printing is a recommended read.

Also look for books authored by Bruce Fraser.


The Crop Factor

Curiously many photographers rarely realise that digital images [from cameras] are not the same aspect ratio as industry standard papers. This is further complicated by differing standards between the US and developed countries, the former still widely foregoing ISO sizes.

A US standard size often used is 8" x 10", which is in 4:5 ratio; 35 mm film is 24 mm x 36 mm, a 2:3 ratio - not really compatible without losing something. Nikon's DX format is - give or take a fraction - 15.7 mm x 23.7 mm, which is essentially 2:3 again.

ISO sizes are based on an ingenious revelation some guys had a long time ago, first put into a standard in 1922 in Germany: With a ratio of 1:square root of 2 every halving of the paper sheet over the long side [parallel to the shorter] results in a sheet with the same aspect ratio!

While DIN A paper is closer to the film formats 2:3 ratio, it is still off**, so you will again lose something. Obviously you should crop beforehand to the desired output ratio.


The 16-Bit-Benefit

The bit depth, OTHO, has a real meaning with regard to your image's quality. In general 8 bit allows for 256 shades per pixel while 16 bit allows for 65,536. Before you get excited, this sounds much better than it is in reality. While it allows for much finer colour changes, no current output device can show 16 bit, no monitor, no printer; actually there is currently no camera capable of producing real 16 bit photos, Nikon's D3-series goes furthest with 14 bit.

One of the reasons most HDR images look so strange, plasticky, otherworldly comes from the fact that our graphic adaptors and printers [or computers] have to remap the original 65,000 shades into 256.

Now, there is a real-life application of 16 bit: If you have an image with very fine gradation, say, a cloudless summer's sun set/rise, or you have to do a lot of processing [levels/curves etc.] you have much more lee-way.


The 16-Bit-Downside

Beside the fact that 16 bit are rather theoretical ATM, these files are double the size of 8-bit photos, which may not be a big problem for storage but it can be for handling. Remember, Photoshop and other programs have to hold the whole file in direct accessible memory - not necessarily RAM alone, but any data held on scratch disc slow down the program a lot. Actually most programs will hold two or even more copies of the file in memory. That can quickly get out of hand.

The reason someone told you to change the mode of the Photoshop-file is that JPEG is 8-bit only, which shouldn't concern you as long as you save in PSD, TIFF, DNG or other RAW formats.





*Obviously I open myself to feminist criticism here. Well, I took 'gal' for three reasons:
1. The original poster seems to be a women.
2. I cannot stand the phrase 'regular guy' anymore.
3. If I had used 'guy' I would equally opened up myself to feminist criticism since it'd imply technical matters are guy matters.

**0.66666666666666666666666666666667 compared to 0.70710678118654752440084436210485 compared to 0.8 [2:3; DIN A; US respectively]
 
72 ppi [pixels per inch] is a remnant of older monitors, particularly Apple, run at that relative resolution; Windows native RR was 96 ppi
72/96 are DPI not PPI. 72 generally was the DPI value for CRT, 96 DPI is standard for LCD.

the RR put into the image metadata is only meaningful if you print without setting a specific output resolution, and only if the program you print from reads it.

The value in the file is never meaningful.

Also look for books authored by Bruce Fraser
I red an article of Bruce Fraser and he makes a lot of confusion between sampling and dithering.
One of the reasons most HDR images look so strange, plasticky, otherworldly comes from the fact that our graphic adaptors and printers [or computers] have to remap the original 65,000 shades into 256.
True, tonemapping must be performed to create an LDR image from the recovered HDR, but good tonemapping produces good and natural images.
 

Dierk Haasis

pro member
72/96 are DPI not PPI. 72 generally was the DPI value for CRT, 96 DPI is standard for LCD.

Are you telling me monitor images are made up of dots not pixels? Dots are used by printers, monitors contain pixels.

The technology behind the monitor - cathode ray or liquid crystal - has nothing to do with the relative resolution; it's purely coincidental that LCDs are not used with 72 ppi anymore, simply because neither the Macintosh system nor Windows uses it anymore. When Windows came up with 96 ppi LCDs weren't even used for monitors, only as very basic dark grey on light grey displays in pocket calculators.

Without anything concrete I find your comment on the late Bruce Fraser irritating. Where did he write what that is confused?

Note my qualifier in the quotation on HDR. I have seen convincing HDRs but very few [to come up with a ratio - 1:100,005]. Since it is not the issue of the original question, let's leave any discussion on the merits [or faults] for another thread.
 
Are you telling me monitor images are made up of dots not pixels?

Every time the pixels are made visible, rendered, they are transformed into dots with a pysical dimension. So, yes any device output DPI not PPI.
Think to contone printers. in this case PPI=DPI, same value for diffrent things.
The technology behind the monitor - cathode ray or liquid crystal - has nothing to do with the relative resolution; it's purely coincidental that LCDs are not used with 72 ppi anymore, simply because neither the Macintosh system nor Windows uses it anymore. When Windows came up with 96 ppi LCDs weren't even used for monitors, only as very basic dark grey on light grey displays in pocket calculators
I don't say that 72 to 96 is correlated to the technology.

Without anything concrete I find your comment on the late Bruce Fraser irritating. Where did he write what that is confused?

I can quote Jeff Skewe:

The actual method of producing the Epson's variable droplet size is an exotic error diffusion algorithm. To the driver, it really doesn't matter whether the resolution is a mathematical divisor.

Sampling vs dithering.

That used to be the common wisdom, however, Bruce Fraser has written that it's far more useful to use the REAL resolution without resampling and letting the PPI fall where it may for a given image size. So, resize without resampling to get the print size you want and let the actual resolution fall where it may–as long as you are in the range of 180-480PPI.

Sampling ? No matter !

However, it's wrong to characterize the Epson driver as doing "resampling" to make it's dots...it doesn't resample. The error diffusion halftoning is far more complex and exotic. As a result, you really DON'T gain any real advantage to actually resample images merely to get to some magic number. In the case of downsampling you are throwing useful pixels away and in the case of upsampling, you are making up pixels out of essentially nothing (just the surrounding pixels).
Sampling vs dithering

See, that's where ya got it wrong...the driver TAKES the image data at it's actual dimension and then runs what is, in effect, an error diffusion process to determine where droplets will and won't be. The driver (at least from the Epson side and this is coming from a guy WAY smarter than me-Parker Plaisted) takes whatever data it's given and runs it through a sieve (the metaphor for the error diffusion), depending on the resolution settings on the driver, the sieve gets larger or smaller openings to create the stochastic halftone that is then broken down into a droplet map to tell the print head when to and when not to squirt some ink
Sampling vs dithering

In my case, with the typical MP (mega pixels-yet ANOTHER unrelated measurement scale) that I shoot and the print sizes I generally want, 360PPI is just about optimal...which is a long way of explaining why I chose to set the image resolution to 360PPI (without much concern over the image size).

How to accept that sampling is important, without naming it.

Bruce Fraser, in his Real World Image Sharpening talked about the kind of resolution you need to NOT see any actual dots when printed...but here again, the problem is human vision isn't measured in PPI or DPI...since human vision is measured in minutes of arc (at about 1.5 minute of arc per line pair) that doesn't translate to dots on a page...what Bruce did was to factor out what the human vision was capable of resolving at various distances (since distance has an impact because of the arc). Bruce figured that a person with 20/20/20 vision in good light could resolve about 355 dots/inch at a distance of 12 inches. Note, the 20/20/20 is a Bruce joke, that equates to a 20 year old with 20/20 vision. Close focus gets poorer the older you get
.

When you see a print, you see DPI, not PPI. DPI value does not change depending on input image size.

The viewing distance is not related to the DPI value, it is related to the driver resampling artifacts.
Long distance is a cure for many problems.

The printing process is a true black magic if you ignore samplig and mix DPI with PPI.


Jeff Skewe is of the school of Bruce Fraser.
 

Dierk Haasis

pro member
First off all, the guy is called Schewe, and I do not equate his opinions with Fraser's. Nevertheless, Jeff Schewe is a very experienced guy, particularly with printing, and he is a practicioner; theory was never his strong side. I am also a bit suspicious you misinterpret or even misunderstand Fraser's, Schewe's, and to a degree Chaney's claims about how an Epson printer [or any other brand] actually goes about to create the illusion of a resolution much higher than its physiology allows. Which is exactly the point I made above on native resolution.

Look at the Epson 2100, which offers 720, 1440 and 2880 dpi. Since the built-in printhead does not have more than 720 active openings, 1440 and 2880 are virtual resolutions - created by dithering! Just like Fraser and Schewe write.

I have no idea why, after taking me on on a purely theoretical matter, you drag in the viewing distance issue. Since you do it, and I wrote about 300 dpi being overkill most of the time, let me add that I am totally with Fraser, Chaney, Schewe and other practitioners: Don't judge prints [or pictures in general] on surreal criteria. Looking at an A3 print at less than arm's length [or even with an 8x loupe] is nonsensical. Full stop.

As for the difference between ppi and dpi, I know I am nitpicking, but if an image is created from individual pixels, like it is on monitors, it's ppi. The result to the viewer between dpi and ppi may be the same, technically there is a difference. It might not be very important on a day-by-day basis but clear definitions are sometimes necessary.

I don't say that 72 to 96 is correlated to the technology.

No, you write it:

72 generally was the DPI value for CRT, 96 DPI is standard for LCD.

That was after I explained how operating systems used these values to determine size on a monitor. There is no inherent relation between LCD-CRT-virtual monitor resolution.


PS: When I equalled 300 dpi with 150 lpi [a printing machine screen frequency] I presumed the standard maximum of 2.0 as the halftone factor; for best results it may be as low as 1.5, making 225 dpi quite sufficient for good prints in normal sizes [up to A2] at normal viewing distances. A look at big posters and outdoor adverts, incl. blow-ups, will tell you that after A2 the relative resolution goes lower and lower with no adverse effect.

PPS: Since digital images have an absolute resolution, that is a fixed number of pixels, neither ppi nor dpi or lpi make sense with reference to them; there's simply no inches involved, only pixels.
 
to create the illusion of a resolution much higher than its physiology allows. Which is exactly the point I made above on native resolution.
This is not to create the illusion of a greater resolution. It’s to create the illusion of colors that are not covered from the inks. I think you know that dithering is an error diffusion problem, the error is the color difference. The proccess has not a spatial relevance but a tonal relevance.

Don't judge prints [or pictures in general] on surreal criteria. Looking at an A3 print at less than arm's length [or even with an 8x loupe] is nonsensical
I agree totally.
My point was to decorrelate DPI value from viewing defects. Defects are originating from poor PPI value not from poor DPI value.
Fraser or Schewe or both made this correlation, I don’t know your thinking.

As for the difference between ppi and dpi, I know I am nitpicking, but if an image is created from individual pixels, like it is on monitors, it's ppi. The result to the viewer between dpi and ppi may be the same, technically there is a difference. It might not be very important on a day-by-day basis but clear definitions are sometimes necessary
.


I agree: the first way to avoid confusioni is to use correct terms.

I think you know that some monitor use 6 bit and dithering, like printers.
As I said contone printers create one dot for each pixel.

Yes tecninically there is a difference. No matter how the image pixels are mapped ( one to one or one to many) when image is rendered DOTS are created.
 
Top