Image Compression

[Chasew]

My day job is in the printing industry. Every day I see someone try to print a something that has been optimized beyond use. Sometimes it’s an image captured from a website, sometimes it’s a client forwarding a PROOF file, that is low res, re-sized, and a different color profile than what the PRINT file was intended to be. If the designer knew their clients were trying to send proof files to print, they’d likely flip out. I know, because I’m one of them. Very frustrating… But it’s the nature of our business, that our clients often know nothing about.

Someone captures a great photo (inevitably with a phone – though that part is somewhat irrelevant, with the advances in phone cameras as of late). This photo is so great, they have to share it. So they text it to someone. That alone has destroyed the image, as far as I’m concerned… But that person inevitably shares it with someone else, or posts it to Facebook, or uploads it to fill in the blank. And then re-saves it (to their phone) and emails it to a print center (from their phone), at who knows what file size/compression. How many rounds of degenerative compression has this thing gone through? Most of it isn’t evident on a 4.7″ screen, but when printed on a typical 8″x10″, or poster for a funeral / trade show / marketing… What once was a great photo is now a mere shadow of it’s former self.

One of my regular clients uploads vendor product photos to a popular file sharing network (optimized). Then links the uploaded image to an email newsletter service, crops/edits the photos in their editor (optimized), and then forwards those post-processed images to me for use in print. When I complain about their low resolution, the response is always something to the effect of “but that’s what they sent us”.

Is it? Is it, really??

Another client has taken screenshots of photos on their phone, and emailed those to me for use in marketing. Literally… battery indicator and all…

People unfamiliar with the industry or the mechanics, have a hard time grasping the general concepts. Somewhat understandably… It reminds me of when my mom scolded me for using dial-up internet back in the day, because she thought I was draining the phone battery. Very silly, but really, how could she know any better? Made perfect sense to her. The internet uses the phone, right? This same level of misunderstanding can often cause problems and delays in our already technical processes. And ultimately, it just bothers me. Most images today are digitally created. Somewhere out there, there’s an original at full res.

Another scenario: I have an edited photo that I want to upload to Facebook. I know Facebook is going to apply their own brew of compression wizardry, but I trust Adobe’s more. So I do my own compression before uploading it. Have I made things better or worse? That’s my question. How do these apps actually compress things. Is there a threshold for file size, where if an image is beyond a certain point, compression is applied? Or does every image get optimized? If that’s the case, then perhaps I was better off uploading the original, unoptimized image, to let Facebook do their own thing. Otherwise, they’ll be auto optimizing an already optimized image. Does that matter? How truly degenerative is multiple optimizations? Depending on how it’s done… Because technically speaking, once something is actually “optimal”, it can’t really get any more optimal…

How do we discern the differences? How do we tell how much quality has been lost? Are we truly working with the “original” images or has there been multiple levels of destruction? How many levels? Does it matter?

I’m interested in other’s thoughts. I might do some research / comparisons to see out how various apps/websites compress images.

[__]

How do these apps actually compress things. Is there a threshold for file size, where if an image is beyond a certain point, compression is applied? Or does every image get optimized?

I’m sure the approach varies wildly across different sites. I imagine filesize is the biggest trigger, followed by pixel dimensions. Beyond that… who knows?

How do we discern the differences? How do we tell how much quality has been lost?

I don’t think it’s possible to know without having the original image to compare to.

I might do some research / comparisons to see out how various apps/websites compress images.

Might be interesting, if you find anything (assuming a broad enough sample for meaningful comparison). Not sure if there would be much of a practical takeaway, though. Are you looking for anything specific, or is this just curiosity born of frustration?

[Alen]

Stick to your process and educate your clients. Don’t try to figure out the level of deconstruction, that’s insane. It’s in their interest to give you the correct files. Explain it to them. It’s your job.

[Author]

Posts