HELP- need image size explained also TIFF vs PSD

Olympus OM-D E-M5 Mark III

04/28/2009 12:19:56 AM · #1

can anyone explain the intricacies of image size or tell me of a site that has a tutorial about it .. FOR DUMMIES .. !!?
i've seen a couple but i'm having a brain seizure trying to understand it ..

when i shoot highest quality jpeg with my canon 40D the resolution is 72 pxls/inch and the document size a whopping 54 x 36 inches ..
when i shoot in RAW the resolution is 240 pixels/inch and the document size is less, being 16.2 x 10.8 inches ..
in both the file size is the same & also the pixel dimensions ie. 3888 x 2592 pxls ..

i am trying to get together images that are print worthy for photographer's direct, my Loupe and alamy ..
they expect photographer to know all this stuff and i dont ..

unfortunately i stopped shooting in RAW recently and got rid of all my RAW files .. that was before i realised i needed them for prints ..
i'll be shooting in RAW from now on ..
i bought Geniune Fractals hoping that i can 'upsize' all my previous work & that they will convert to large prints and still retain the quality ..
i've 'upsized' a couple of images using Geniune Fractals, but blown right up cant see a difference .. but that's probably for another day .. ??

BUT .. lets say i've done all my processing on a file and want to save it as a jpeg for print & i want it to be 54 x 36 inches but it started out as 16.2 x 10.8 inches because it was processed from a RAW file ..
if i increase the document size or pixel dimensions wouldnt the resulting image lose quality if i didnt use any software like Geniune Fractals . ?

also if i started with a jpeg that was already 54 x 36 document size yet the resolution was only 72 pixels/inch would that convert to a print of that size seeing it has less pixels/inch and the pixels might become visible .. or can i actually increase the resolution and retain quality ... i actually tried that yesterday and the image became unmanageably huge ..

its hard enough even trying to phrase this question ..
i'm having trouble understanding pixel dimensions, document size and resolution and how they relate to each other ..
also when you increase any of them then wouldnt the size of the file become too large ?
i mean i understand that if the resolution is 72 pixels/inch then there are 72 pixels in each square inch of the image ..
and if the pixel dimensions are 3888 x 2592 then there are 3888 pixels along the width and 2592 for the height ..
oh ... i just realised how that relates to document size .. of course the document size will be smaller if there are 240 pixels/inch rather than 72 pixels/inch ..
but its so complicated ..
sorry to go on, but i really want to work this out ..

ALSO ....
i've heard ppl talk about TIFF format before and sort'v shrugged it off assuming that working in PSD was ok for anything i ever wanted to do ..
i'm starting to realise that this may not be so ..
can anyone tell me the advantages of working in TIFF instead of PSD ..

also .. my working space is sRGB IEC61966-2.1
when i open a tiff file that i've saved from a raw file or some other files i'm asked if i would like to .
- use the embedded profile (instead of the working space)
- convert document's colours to the working space
- discard the embedded profile (dont colour manage)
any opinions on that ?
i started using the embedded profile but when i saved as a jpeg and put online the colours were a lot less vivid ..
since then i've been converting documents colours to the working space ..

I USE PHOTOSHOP CS3 .

thankyou anyone who can help .. :)

Message edited by author 2009-04-28 17:54:46.

Yo_Spiff

Olympus 12-100mm f/4.0 M.Zuiko ED IS PRO

04/28/2009 12:58:34 AM · #2

A subject I have become familiar with, both on and off the job! I spent 6 back and forth e-mails on this subject with a coworker the other day. It's still tough to explain, however. One day I will hit on the explanation that works first time for everyone. That means you are my latest guinea pig!

DPI is a measurement of the pixel (or dot) density of a printed page (or how finely it is being scanned). The actual size of an image is measured in pixels and for a given image, that pixel count remains the same regardless of the assumed page size and DPI setting.

DPI is only an abstraction and makes no difference until a piece of paper comes into the scene. (whther you are printing it or scanning it) Some software will make a default assumption of the DPI setting, and perhaps embed that assumed setting into the file, but it does not change the actual image, it only changes the default physical size at which it will print.

Here are a couple of identical images I have done. The ONLY difference is the DPI encoded in the file, otherwise they are the identical image

600 DPI:

184 DPI:

If you printed these at the DPI encoded in the files, the 184 DPI version would print physically larger, but with the pixels spread out over a wider area. The 600 DPI version would be physically smaller, but would still be composed of the same number of pixels, just packed tighter.

The printed document size is a function of both the size in pixels and the DPI it is being printed at. For example, lets use a 720 pixel challenge entry:

At 300 DPI it prints at 2.4 inches 720 pixels(dots) divided by 300dpi =2.3 inches

At 72 DPI it prints at 10 inches 720 pixels (dots) divided by 72dpi = 10 inches

Here is an article on the subject that I found a while back Maybe this can help clarify (or confuse) it some more.

As far as your workflow, PSD is better for files in work, because it preserves your layers and other Photoshop specific information about the image. TIFF is a more standardized format however, and is preferable as a format to save the final results. In 50 years someone may not be able to open a proprietary PSD file, but the more standardized TIFF will likely be readable. TIFF is usually the format recommended for digitally saving images for genealogy work. Both TIFF and PSD are formats that do NOT use lossy compression like JPEG, so you so not lose any image quality when saving. Any compression they may use is lossless.

Message edited by author 2009-04-28 01:18:42.

Ken

Canon EOS-7D

Canon EF 24-70mm f/2.8L USM

04/28/2009 07:36:48 AM · #3

I think the explanation

Yo_Spiff gives is really good and makes sense to me. I had a general understanding of it but now I'm clear.

But Roz, why stop shooting raw and get rid of your raw files? I always save my raw files for archival purposes and the hope that one day my processing gets better and I can go and re-edit from the original, high quality source.

Sony 24-70mm F4 Vario-Tessar T* FE OSS

04/28/2009 05:57:42 PM · #4

Originally posted by Ken:

I think the explanation Yo_Spiff gives is really good and makes sense to me. I had a general understanding of it but now I'm clear.

But Roz, why stop shooting raw and get rid of your raw files? I always save my raw files for archival purposes and the hope that one day my processing gets better and I can go and re-edit from the original, high quality source.

thankyou so much steve

Yo_Spiff you've given me a lot to think about .. :)
hey

ken .. sorry i wasnt clear .. i've edited my question .. but i stopped shooting in raw just because they take up so much room on hard drives and i didnt realise i'd be needing them if i wanted to do prints of any quality and size ..
i'm going to be shooting in raw from now on .. !!

does anyone have any more info ... just hearing it from different sources can be good ..
& what about resizing and losing quality .. if i've processed an image that started as a raw file and the document size is only say 10x8" .. can i just change that to 24x36 or wotever and still have a quality product .. ???

& wot about the working space .. :)
"my working space is sRGB IEC61966-2.1
when i open a tiff file that i've saved from a raw file or some other files i'm asked if i would like to .
- use the embedded profile (instead of the working space)
- convert document's colours to the working space
- discard the embedded profile (dont colour manage)
any opinions on that ?
i started using the embedded profile but when i saved as a jpeg and put online the colours were a lot less vivid ..
since then i've been converting documents colours to the working space .. "

Message edited by author 2009-04-28 18:03:58.

Bear_Music

Samsung Galaxy S24 Ultra

04/28/2009 06:12:15 PM · #5

Originally posted by roz:

& wot about the working space .. :)
"my working space is sRGB IEC61966-2.1
when i open a tiff file that i've saved from a raw file or some other files i'm asked if i would like to .
- use the embedded profile (instead of the working space)
- convert document's colours to the working space
- discard the embedded profile (dont colour manage)
any opinions on that ?
i started using the embedded profile but when i saved as a jpeg and put online the colours were a lot less vivid ..
since then i've been converting documents colours to the working space .. "

Adobe RGB is a "larger" color space than sRGB: in other words, it stores much more color information. Your camera should be set to AdobeRGB: when processing for DPC, at some point in the workflow you need to convert the image to sRGB (which is the standard for web display), but it's good to have the larger color gamut available for critical, real-world tasks like printing or photo reproduction. When shooting RAW, it actually doesn't matter what color space the camera is defaulting to, since you can assign any color space yopu want to the RAW image, but if you're shooting JPG then the assigned color space is what you're stuck with, information-wise: in other words, you can go *down* from AdobeRGB to sRGB in post, but if you try to go *up* it's not acftually adding back lost color information.

R.

GeneralE

Canon PowerShot S3 IS

Canon 12x IS USM (S3 IS built-in Lens) -- 36-432mm (35mm EQ)

04/28/2009 06:14:18 PM · #6

Originally posted by roz:

does anyone have any more info ... just hearing it from different sources can be good ..
& what about resizing and losing quality .. if i've processed an image that started as a raw file and the document size is only say 10x8" .. can i just change that to 24x36 or wotever and still have a quality product .. ???

If you resize without resampling, you have the same number of pixels, so the integrity of the underlying image is unaffected. As a rough guide, as long as you have 150pixels/inch you should get a good print output, so a 24x36 inch image would need 3600x5400 pixels (24x150=3600, 36x150=5400).

Once you get into resampling the image, the results are largely dependent on the quality and amount of detail present in the image.

A Quick Visual Guide To Resampling

Message edited by author 2009-04-28 18:16:11.

yanko

Canon EOS-5D Mark II

Carl Zeiss 50mm ZE f/1.4 Planar T for Canon

04/28/2009 06:49:48 PM · #7

Originally posted by roz:

can anyone tell me the advantages of working in TIFF instead of PSD ..

If you're creating layers in Photoshop and want to preserve those layers for editing latter then it's better to save as a PSD file. TIFFs can save layers, but from my experience the file sizes end up much larger than the PSDs. The real benefit to TIFFs is when it is compared to JPGs. I usually will keep a PSD file for my edits and then when I am done I flatten the file and save as a TIFF since it has lossless compression or no compression options unlike saving as JPG which even at the highest compression setting still losses some quality.

Message edited by author 2009-04-28 18:52:43.

Canon EF 24-70mm f/2.8L USM

04/28/2009 07:57:57 PM · #8

Originally posted by Bear_Music:

Originally posted by roz:

& wot about the working space .. :)
"my working space is sRGB IEC61966-2.1
when i open a tiff file that i've saved from a raw file or some other files i'm asked if i would like to .
- use the embedded profile (instead of the working space)
- convert document's colours to the working space
- discard the embedded profile (dont colour manage)
any opinions on that ?
i started using the embedded profile but when i saved as a jpeg and put online the colours were a lot less vivid ..
since then i've been converting documents colours to the working space .. "

Adobe RGB is a "larger" color space than sRGB: in other words, it stores much more color information. Your camera should be set to AdobeRGB: when processing for DPC, at some point in the workflow you need to convert the image to sRGB (which is the standard for web display), but it's good to have the larger color gamut available for critical, real-world tasks like printing or photo reproduction. When shooting RAW, it actually doesn't matter what color space the camera is defaulting to, since you can assign any color space yopu want to the RAW image, but if you're shooting JPG then the assigned color space is what you're stuck with, information-wise: in other words, you can go *down* from AdobeRGB to sRGB in post, but if you try to go *up* it's not acftually adding back lost color information.

R.

thankyou robert

Bear_Music..
i did change to adobeRGB on the camera a while ago .. but everytime i brought a file into photoshop it asked me wot i wanted to do with it .. so i changed back to sRGB ..
also i changed the working space some time ago from sRGB IEC61966-2.1 to adobe RGB .. .. it was a pain, if i wanted to open a few to check them out and cull them, having to ok them one at a time ..
but just this minute i changed it back to adobe RGB on the camera !!
but do i .. ??
- use the embedded profile (instead of the working space)
- convert document's colours to the working space
- discard the embedded profile (dont colour manage)
it always asked me this when i bring in a file that has been 'processed' in my raw program ..

i changed my working space back to sRGB IEC61966-2.1 but should i change it back to adobe RGB .. i've forgotten where tho .. !!
i dont understand wot to do when i'm asked if i want to embedded, convert or discard .. like if i convert from adobe RGB to sRGB IE61966-2.1 would i then be losing colour information ..? ..:)

thanks paul

GeneralE.. i've got a headache .!!. think i'll go to the gym and get some oxygen in my brain and tackle the resampling link when i get back .
and thanks richard

yanko.. i worked on a TIFF file with layers for the first time yesterday, and the file size seemed to be wot i'd expect if i'd been working in a PSD format .. but saving as a TIFF after removing the layers is a great piece of advice ... :)

Ken

Canon EOS-7D

04/28/2009 08:26:38 PM · #9

Roz, take the time to really learn how to use Genuine Fractals, it can do a great job on resizing. You can great results reducing to web size and sharpening in one step.

04/29/2009 03:22:24 AM · #10

what are the advantages of these options and in what situation should i use them ?
- use the embedded profile (instead of the working space)
- convert document's colours to the working space
- discard the embedded profile (dont colour manage)

regarding resizing ..
if i have an image where the resolution is 240 pixels/inch and i scale that layer larger what happens to the pixels ..
are they expanded and if they are then wouldnt there be less in a square inch .?.
and if there are less pixels/inch after upscaling the layer, it wouldnt be reflected in the image size window .. ?
i've got another headache .. :)

& thanks

ken ..

Message edited by author 2009-04-29 03:25:30.

Canon EF 24-70mm f/2.8L USM

04/29/2009 05:53:22 PM · #11

can anyone out there help me please ...
what are the advantages of these options and in what situation should i use them ?
- use the embedded profile (instead of the working space)
- convert document's colours to the working space
- discard the embedded profile (dont colour manage)
i've noticed that if i discard the embedded profile the colours become very dull ... ?

regarding resizing ..
if i have an image where the resolution is 240 pixels/inch and i scale that layer larger what happens to the pixels ..
are they expanded and if they are then wouldnt there be less in a square inch .?.

Ken

Canon EOS-7D

04/29/2009 06:12:07 PM · #12

Originally posted by roz:

regarding resizing ..
if i have an image where the resolution is 240 pixels/inch and i scale that layer larger what happens to the pixels ..
are they expanded and if they are then wouldnt there be less in a square inch .?.

I believe they are interpolated - the program will take information it has and fills in the blank areas.

On a side note, have you ever seen those blank verticals in your histogram when you do curves or levels? That's missing information that Photoshop fills in. Take a 16 bit from raw and a jpg and do the same adjustments for a demonstration of why raw is better.

yanko

Canon EOS-5D Mark II

Carl Zeiss 50mm ZE f/1.4 Planar T for Canon

04/29/2009 06:27:29 PM · #13

Originally posted by roz:

You should keep the embedded profile. If you ever need to change it you can do so via the menu option: EDIT>CONVERT PROFILE. As Bear pointed out earlier, AdobeRGB is a larger color space than sRGB so it makes sense to set the AdobeRGB color space in your camera and then after you're done editing your file in photoshop and want to save off a web version just convert the profile from AdobeRGB to sRGB via the aforementioned menu option. Doing it this way preserves the best possible color and tone. What it does is remaps the values over to the new color space and it does a much better job of it than when you just discard the profile hence why it looks dull when you choose that third option.

Message edited by author 2009-04-29 18:28:39.