Resizing a photo for photo contest?

Thinking about entering a small local photo contest, and I came to this requirement:

Photographs submitted for entry must meet the following requirements: (i) photographs must not
exceed 700 KB at 72 ppi (ideally images should be between 1000 and 1800 pixels on the longest side)...

When I try to resize my image in photoshop, I select 72 ppi (original file says 300 ppi) and the file is reduced to 441 x 661 and 854KB (down from 1837 x 2756 and 14.5MB). I must be doing something wrong. I am still too large for the 700 KB limit yet nowhere near the "1000-1800 pixels on the longest side" that the contest recommends. What should I do?

Thanks!

Ignore DPI, that has nothing at all to do with your file. Just worry about the dimensions and final size requirements. Any contest that puts that as a constraint doesn't have a clue what it means.

Crop/resize to get 1800 on the long end, then save as JPG, at a compression that meets the file size. Sadly their constraint there too is very poor, 800kb is not much or wont be high quality depending on the subject material.

Go straight to "save for web and devices"

There you can fiddle with the size and compression to get the best output. It will tell you an accurate file size before you export it.

Left Handed Brisket wrote in post #18605791
Go straight to "save for web and devices"

There you can fiddle with the size and compression to get the best output. It will tell you an accurate file size before you export it.

And if you don't like the final result you may want to revisit and sharpening and noise reduction.

Also remember that in Ps with the exception of the Save for web, or the tool that has replaced it in later versions, the image size that you see is for the uncompressed image. So for example in the image resize dialogue above, the Image Size refers to the amount of memory the uncompressed RGB data is filling, based on the pixel dimensions shown. For 1837×2756 you get 5062772 pixels, and you need three bytes per pixel for 8 bit RGB colour, so that is a total of 15188316 Bytes, so when you convert to MB that gives you 14.484 MB. So for the image at 1800×1200px that will give an uncompressed image size of 6.2MB.

Although 700 KB is quite tight for file size, it shouldn't be too hard to meet. The old POTN limit of 1024 Px long edge and 150 KB was usually pretty easy to get to for most images. You could even generally leave the EXIF data embedded in the file and Q80 was often possible. IMO after considerable testing you cannot see the difference between Q80 and Q100, and it's almost not even measurable. You should be able to go down as far as Q60 without appreciable loss of quality. If it were me and I couldn't make 700KB in this case I would strip out all the EXIF data other than your copyright notice first before reducing the quality below Q80. Given that you have a lot of plain blue sky in the image I would think that you would get this down to 1800×1200 with EXIF at Q80 with a file size of under 200KB, it looks to be a quite compressible image. Oh and even images that are going to be printed I only save at Q80 when sending them to the lab.

Alan

The old POTN images also didn't look very good. Often people were unhappy with what they attached to posts vs higher res more detailed images from external sources. 700kb isn't going to be great for every shot though, something with lots of detail could suffer, or something with smooth color gradients might show the compression from JPG 12 down to JPG 9 for example. Just depends on the photo contents, I suppose. Probably all fine for a publication photo I guess, but it seems like a silly constraint these days. I could see a limit of 1.5M.

EDIT: I took a shot of a flamingo and tried it out, and I didn't see too much of an impact before and after, with the feather detail.

I recently discovered and tried JPEGmini. Great for reducing file size with no discernible impact on quality. Save your Ps file as a jpg 1800px on the long side at a resolution that gets you a jpg about 1.5Mb, then run it through JPGmini and it should come out less than 700kb. If you download the 'lite' version of JPGmini, it's free and you can run 20 photos per day through it forever. I've found it so useful for photos I email that I'll probably stump up the cost of the full version, to not be limited to 20/day.

Thanks for all the responses, everyone!

Here is what I did. In PS CC I used the keyboard shortcut CTRL + ALT + SHIFT + S (on Windows) to save for the web. I resized to 1800 on the longest side and played with the quality slider until I was under 700kb. Usually that was between 75-85.

I appreciate all the comments

Do a mild sharpening after the resize to put detail back in before saving out. It will help a bit.

One thing to be aware of when using the SfW or Lr is that Adobe still only have 13 actual levels of JPEG compression. So Q70 to 77 equates to level 9 and 78 to 85 are level 10. To get to 11 you need to go to 86, the highest level is 93 upwards. Remember that for both scales they go from 0, so 0 to 12 gives you 13 levels and 0 to 100 gives you 101 levels.

So although I use Q80 I could just as easy use any other value between 77 and 85 for the same identical result. I have no idea why Adobe would go over to using a 0-100 scale, with only 13 levels of compression, especially when they still use the 0-12 scale widely too.

When it comes to the actual compression effects level 10 seems to offer just about the ideal. I did some comparisons using images saved at all 13 compression levels, plus an 8 bit TIFF file. Even when you save a file at the maximum quality as a JPEG, with no compression, you still get conversion artifacts because the JPEG file does not store the data as RGB triplets, it uses luminance and chrominance values in a similar way to how analogue colour TV works.

In my tests as well as judging the image quality by eye I did numerical evaluations by putting all of the images as layers in a Ps image. By using the difference blend mode you can actually measure the difference between any two versions of the image. When comparing both the Level 10 image, and the level 12 to the original TIFF file version the size and distribution of the differences is pretty much identical, with a large majority having 0 difference. The differences that do exist are generally within the range ±2 with maxima being about ±6. Oddly if you compare the level 10 with the level 12 they have a higher range of differences than when you compare them with the TIFF.

The great advantage with using level 10 is that you get somewhere between a 60% and 40% reduction in file size on disk, compared to the uncompressed image, with minimal measurable degradation, and absolutely no visible degradation at least as far as I can see. As I have said before I now never use higher than level 10 when saving files as JPEG, even when sending the files out to the lab for printing.

Alan

Alerases wrote in post #19267617
When printing photos, it is essential to use high-quality materials. The highest quality photo paper will look like the real thing when viewed 100% on your monitor.

How is that relevant to this discussion (which ended 3 years ago)?

The 'dpi' number is a FALSEHOOD misunderstood by the print competition folds...one can have a 4000x3000 count with '72 dpi' or '300 dpi' embedded in the EXIF data with ZERO VISIBLE DIFFERENCE in the viewing of that image on a computer screen.
That spec ONLY really matters when you send the image to an offset press, because it controls the 'screen density' (dots per inch) used to reproduce the image on the printed page! 75dpi is a low quality offset print image of a newspaper, 300dpi would be a high quality image seen in National Geographic, for example. DPI in the EXIF of a JOG means nothing when you make an inkjet print!

So ask your software for

1. 75dpi
   
2. output for up to 24" print
   

Wilt wrote in post #19267688
<snip>
The thread may be 3 years old, but the issue is still misunderstood even now! Ergo my post.

I can't believe the number of people who still talk about digital images in a digital context and obsess over dpi. It's so strange...

bseitz234 wrote in post #19267694
I can't believe the number of people who still talk about digital images in a digital context and obsess over dpi. It's so strange...

Not really strange. The industry has CAUSED the confusion!

• The software often confuses 'dpi' vs. 'ppi'
  
• The professional print making services has long confused 'dpi' vs. 'ppi'
  
• Those writing requirements for contest submission still confused 'dpi' vs. 'ppi'
  

I guess 3 years later some still need to continue the discussion...