>>but why does an original 300x300 image at 300dpi look fine but the same image >>300x300 at 72 dpi look crappy ? i’m talking on-screen.”
>>I don’t know what that is in reference too, but those two images would be exatly
>>the same on screen.
okay i'm saying dpi becasue that's what my photoshop calls it. but it's really ppi.
and 300 ppi is way more information than 72 ppi. the pixel dimensions (300x300) are "the same" but one is native to the screen format (72dpi) and the other is forcing interpolation (300dpi). broadcast tv is the functional equivalent of 72 ppi.
so that's why dpi remains relevant to me, even though my work output remains
"on screen". my head is spinning. all i know is i have to deal with this everyday
and the results are what matters. i get results by keeping dpi in the mix,
even though technically it's really ppi and even though the work never goes to print.
i certainly respect your expertise and i hereby officially relinquish any such claim.
perhaps i've been laboring under an elaborate and unecessary work-around.
but your explanation, while "technically"correct, still doesn't explain what i have to deal with, and what i suspect was the reason patty's images were crunchy on her site.
This is a subject that makes me cry and reminds me of how bad I am with numbers. I had both my boss and a co-worker sit me down not that long ago and help me understand the difference. I still don't fully grasp it. I live in a dpi print based world. But apparently the web doesn't care about dpi, only pixels.
DPI and PPI ExplainedDisplay, Printing, dpi and ppiUnderstanding ResolutionThe Difference Between DPI and PPI
Darrel, thank you. The light just came on for me when you wrote that it's on screen as a pixel.
i think a yet-unstated part of the issue is that the monitors we use have a pixel per inch.
my understanding is that NTSC broadcast monitors are the functional equivalent of 72dpi.
... and now that we have HD television all those old commercials are going to look like interpolated crap. As well as opening yet another can of worms.
The problem with both video and web is that there is no absolute like an inch to go by. It is like units in type. A typical font is drawn at either 1000 or 2048 units. The same number of units exist no matter what the type size is. The output device determines how much of this information it can output but it can't change the number of units. A pixel is like a unit which can be bigger or smaller depending on the device it is viewed on. I can take my single monitor and change its resolution to several pixel depths. This will enlarge or shrink the pixel as well as cut off or add more image in the viewable area. What you like to call an inch is actually a variable that follows pixel resolution of the viewing device.
"NTSC video is the equivalent of 72dpi"
yea, that's technically incorrect. NTSC is simply a pixel dimention. It's not equivalent to any DPI.
What might be confusing is that a lot of DTP software defaults to assuming 72dpi = 100% on screen. This was only technically true when the first Macs came out (when monitors were roughly only 72ppi). It hasn't been true since, but software still uses it.
For instance, if you set an image to 72dpi in Photoshop and then tell photoshop to display it on screen at 100% you'll get actual size in pixels. Changing the DPI in photoshop won't affect the pixel dimensions of the software, but will change the 100% view in photoshop. Perhaps video editing software adopted that long ago and never let go.
>>“NTSC video is the equivalent of 72dpi”
okay at this point i just have say i mean FUNCTIONAL and not TECHNICAL.
there's a big difference.
i still have to work according to this 72dpi maxim. obviously i'm not an engineer.
the engineering explanation makes my head spin but the 72dpi assumption
gets my job done ! that's all i can say.
"my understanding is that NTSC broadcast monitors are the functional equivalent of 72dpi."
NTCS = 484 vertical lines. I know not all of those lines are availble on screen, but for this, we'll assume they are.
For the screen to truly be 72dpi, it'd have to be 484/72 = 6.72" high. Any screen larger than that would be less than 72dpi and any screen smaller than that would be more than 72dpi. Same pixels. Just less or more per inch.
"and 300 ppi is way more information than 72 ppi"
Not at all. It's just a different setting in the images meta information. What matters is the number of pixels. A 1000dpi 300x300 image and a 72dpi 300x300 image both have exactly the same amount of image data: 90,000 pixels.
EDIT: I can see the confusion. Your above statement is correct, 300ppi is more information than 72ppi. What you are confused about, though, is that the image on the screen doesn't have any ppi settings. It's simply a bumch of pixels. PPI is a measurment of screen pixel density...not a measurment of anything displayed on it. DPI is a measurement of print dot density...not a measurement of the electronic files that are being printed.
RESSSSUMAYYYYZZZZzzz z z z . . .
"i still have to work according to this 72dpi maxim."
Is this a setting in your video editing software? Which software is this? I find it to be an odd setting and now I'm curious... ;o)
yes apologies to the moderator. we really did get diverted. somebody started talking about patty's web images. may i continue ...or can we dissect this thread into two ?
what is the appropriate protocol for a big new discussion on a thread in which the original question was addressed and then diverted ?
>>Is this a setting in your video editing software?
no, it only exists in photoshop. i create graphics and import them into aftereffects.
almost always it's desirable to work at very high rez (600 dpi or higher) so that aftereffects can pan & scan or reduce as necessary. occasionally i have to work at a lower rez to avoid artifacts created in the video document such as unintentional moire patterns or banding in hi-rez gradations viewd in low-rez media. i always have to preview work on an NTSC monitor to see what it really looks like. the engineers i worked with back in the 90's consistently refered to NTSC as the functional eqivalent of 72dpi. perhaps they were just being kind and sparing me the more complex explanation.
sometimes i have to take frame grabs from video back into photoshop for a promotional publication. they always come in at 72 dpi at the video pixel dimensions.
screen snapshots on my mac also import into photoshop at 72 dpi with dimensions maintained.
when i was prepping stills for my website, my web builders requested that i supply the images at the correct pixel dimensions at 72dpi (photoshop) to avoid any forced interpolation by resizing. i tested a few and saw noticable differences online, similar to the artifacts on patty's site that started this diversion, hence my advice that she follow the same guidelines.
Chris, I'm with ya; everything I'd learned led me to believe that it was all about dpi (72 v. 300, etc.), but Darrel's feedback pushed me to test what he was saying, and the experiment proved he was right. I still can't quite wrap my head around it, but the experiment means I don't really need to, the proof is there. I still know I can't use any of these graphics for print (unless I want to print the image really small), but this does mean I no longer have to resize my graphics (for web) in 2 stages (resample to 72dpi then resize the dimensions). Thanks, Darrel, for driving this one home.
Stephen (and Eddie), more than a hijack I think this thread has now totally changed direction. I apologize as I think I forced it down this other path (see below for link to new thread for this topic). Hopefully Eddie got some good feedback before the thing went screwy.
I think Stephen was hinting that we should start a new thread. Jason, would you do the honors? Make sure to include a link back to this thread. :^) Thanks!
Sure Tiff, here's the new thread:
"Is dpi relevant to web graphics?"
Columbia College (Chicago) has a handy pdf on building resumes.
I'm working on polishing mine as well. If anyone else has a link to their own resume or ones they admire that would be appreciated. Really!
If anyone here is still interested in discussing resumes, the links to the AIGA article given earlier didn't work. I've found a new link to the article:
Ten common mistakes in resumes and cover letters