size and porportion help

[rachelp61]

Hello everyone ,
I hope someone here can help me with this problem. I have been assigned to place a row of people standing side by side of each other . They were photgraphed Individually , with some posing sideways , some standing face forward with arms folded..etc. i hope you get the general idea.Since they were photographed separately and need to be copied and pasted onto one main file...I have no reference of size ( meaning who is taller than who ) or how to adjust . The files I have received come from two different photographers from two different photography studios and also the files ,some are 8 x 10 300 dpi .tiffs and others are 20 x 30 72 dpi jpegs . Is there something in photoshop that will help me with this because I have exhausted all of my thoughts and ideas. Any help would so be appreciated.
Thank you so much in advance !!

 

[RTContent]

What exactly is the issue? Is it that you want to get an idea of how tall some people are compared to others? Or is it that you want to just change the dimensions of the individuals as you see fit? If it's the latter then using the transform tool while constraining proportions will do the trick (Ctrl/Cmd + T, and hold shift while changing).

 

[ALB68]

First thing I would do would be to resize all of them to the same resolution. If your trying to move and/or copy images from photos of two different resolutions will give you results that are definitely out of proportion. May be other things, but without seeing the images it would just be a guess.

 

[Tom Mann]

ALB, if, by "resolution", you mean the ppi or dpi settings for the image, forget about changing it. Doing so is, at minimum, a waste of time, and at worst, a way to blur the images. The ppi/dpi parameter is essentially unused in post processing. It only comes into play if you want to print on a printer that happens to recognize that parameter. All that really matters in post processing is that each of the component images has a reasonably large number of pixels in each direction when you crop each down to just the person.

In addition, be aware that each time you scale a layer (eg, using the cntrl-T, transform tool), you lose some information (ie, sharpness). Because of this, you want to take all measures to avoid a series of sequential re-sizing steps. Instead, I would suggest to the OP that he do more or less what RTContent suggested, but with an additional suggestion: re-size the person in each of the larger images (in pixel dimensions) down to an appropriate size relative to the person in the image with the smallest pixel dimensions. Doing this ensures that you are never up-rez'ing any image, only down-rez'ing the component images.

@ the OP: The much more difficult problem you will face is that among all the component images, the lighting certainly won't be consistent, nor will the contrast, color balance, perspective (ie, focal length of lens used and angle of the camera relative to the person, ie, slightly below or above the subject, from one side or the other, etc.).

To be honest, if you are not quite experienced in PS, OR, if you know that the end user is very accepting of an odd-looking result. I would suggest that you not accept this assignment. Without a reasonable amount of experience, the result is likely going to be a people version of the dreaded "ransom note effect" ( http://en.wikipedia.org/wiki/Ransom_note_effect )

Tom M

 

[ALB68]

Maybe my terminology isn't correct and I don't know much..compared to you of course. I do know if you have an image that is say 2200 pixels wide at 72 ppi and you have an image that is 9000 pixels wide at 300ppi, no way can you copy and paste from the 300ppi image into the 72. If you change the 300 ppi in Photoshop resize to 72 ppi then you can do it. No resample. Vice versa, a copy from the 72 copied into the 300 will be mucch smaller. All I'm saying is the images have to be compatible to one another in this regard. Not taking quality and other issues into account.

Good advice on not taking the assignment, however, sounds like it's part of her job to get it done. She may not have a choice.

 

[Tom Mann]

Hi ALB -

re your statement, "I do know if you have an image that is say 2200 pixels wide at 72 ppi and you have an image that is 9000 pixels wide at 300ppi, no way can you copy and paste from the 300ppi image into the 72. If you change the 300 ppi in Photoshop resize to 72 ppi then you can do it. No resample. Vice versa, a copy from the 72 copied into the 300 will be mucch smaller. All I'm saying is the images have to be compatible to one another in this regard...."

To check which one of us is correct in this matter, I did a little demonstration experiment. First, I constructed two separate photoshop files. The first is a dark blue flower shape on a light blue background that is 220 pixels square at 72 ppi. The second is a series of red bars on a square, muted red background that is 900 pixels square at 300 ppi.

The first thing I tried was to make a copy of the red image which I then pasted (at ~50% opacity) on the blue image. The result is shown in the attached screen grab. It pasted perfectly, but, of course, because it has more pixels, it extended beyond the bounds of the blue image. You can, of course, use the handles to make the red image similar in size to the blue image.

 

[Tom Mann]

The next experiment I did was exactly the same as the previous one, EXCEPT that, as per your post, I changed the ppi setting of the red image from 300 to 72 (without resampling) before I did the cut and paste operation. The result (attached below) is EXACTLY the same as the first experiment.

This confirms that, as I said in my earlier post, changing the dpi/ppi setting without resampling has absolutely no effect when you are cutting and pasting images, ie, making a composite image using PS. All that matters is the pixel dimensions.

 

[ibclare]

My additional statement which doesn't address most of your question, but has to do with changing images is this. One of my favorite improvements in PS is the smart object. Anytime I am unsure of whether I'll be resizing an object, I convert to smart object. I've never been sorry. You can always rasterize it to a regular layer when you're satisfied, but in the meantime, you can go larger or smaller and back again without losing your original image as it's sized.

As for how tall people are compared to each other, you may have to guess based on an average. I can explain more of what I mean if you want. When you're done, if there is a gross error or some gross rudeness? tell them to provide you with each person's height, lol.

 

[Tom Mann]

The final demonstration that I performed was exactly the same as the previous one, except that when I changed the red image from 300 to 72 ppi, I used resampling. Exactly as expected, this reduced the pixel dimension of the red image to 216 pixels square. Of course, now when you do the cut and paste operation, the bounds of the down-rez'ed red image no longer extend beyond the bounds of the blue image. (see attached).

While this procedure gives you the convenience of not having to adjust the size after pasting to see the entire red image:

(a) the OP is going to have to do a second re-size operation to fine-tune the relative sizes of the component images to make the people have correct relative sizes; and,

(b) because of the loss of resolution that occurs each time you down-rez (ie, with resampling), I recommended to the OP that when he is in the process of resizing all of his component images to make the people have reasonable sizes relative to each other I recommended that he not change the pixel dimensions either with or without resampling. He simply should do it once to get the people to have reasonable relative sizes.

I hope this is clear.

Best regards,

Tom

 

[Tom Mann]

Hi Clare -

Using smart objects is an EXCELLENT suggestion. Thanks.

With respect to using an average size approach, I would again caution that one never wants to be in a position where you are forced to up-rez an image. This is why I said in an earlier post:

"...re-size the person in each of the larger images (in pixel dimensions) down to an appropriate size relative to the person in the image with the smallest pixel dimensions. Doing this ensures that you are never up-rez'ing any image, only down-rez'ing the component images. ..."

Cheers,

Tom

 

[rachelp61]

I can not thank everyone enough for your input on this topic !! I have been using photoshop for some time now but this time, I was simply not sure how to handle this in a simple manner . This has helped me sooooo much !!!
Thank you Guru's , you rock .

 

[rachelp61]

This method was what I was using but questioned myself..... this works and again thank you all for your insight !!!!

 

[rachelp61]

Thank you so much for your help !!

 

[rachelp61]

It is not a choice, you are correct. Thank you . ALB Laser so much for your input as well. This was my issue, different sizes and not sure what the best way to handle this . You guys are awesome !