Image compression - CS 5.5

[kaiserwilhelm]

For over six years now, we have used basically the same Export of pdf out of INDD. Sure, over time, we went up to 1.7 or 1.8 and stopped flattening. However, the "guts" concerning fonts and images have remained the same.
I guess I am a tad old school and thought that the more DPI, the better. Thus, we had ZERO compression.

I go into a Trueflow box as appe and come back out that way as well. I don't think that has anything to do with my question, but I thought it fair to put it in.

I am "just now" noticing that I can leave the dpi the same (do not sample) box, but at the same time have the file go to jpeg.
This made an 11 meg pdf go to 1.3
We are seeing over a gig of pdfs per day with many of them transferring to our sister company on the east coast. Transferring a gig is getting to be a pain.
I guess my question is this. Has anyone had any issues with clicking on the jpeg button for CMYK and Grayscale pics?

 

[gordo]

For over six years now, we have used basically the same Export of pdf out of INDD. [SNIP]
I guess I am a tad old school and thought that the more DPI, the better. Thus, we had ZERO compression.
[SNIP]
I am "just now" noticing that I can leave the dpi the same (do not sample) box, but at the same time have the file go to jpeg.
This made an 11 meg pdf go to 1.3
[SNIP]
I guess my question is this. Has anyone had any issues with clicking on the jpeg button for CMYK and Grayscale pics?

For clarity, there's "compression" and then there's "resampling" they can both make the file size smaller but do so in different ways. If you compress the image using JPEG the dpi remains the same. JPEG compression makes the file size smaller by chunking similar image pixels that have slightly different color values into groups of pixels with the same color value. If you use a high quality/low compression setting for creating the JPEGs it is very unlikely that you'll see any degradation in the final press work. That is because the size of the artifacts created by JPEG compression are too small for your halftone screening to resolve.

There's a posting on my blog, with samples, of how different JPEG scenarios play out. Click on this link: Quality In Print: JPEG images for print production - the facts

best, gordo

 

[kaiserwilhelm]

Thank you Gordo!

 

[kaiserwilhelm]

So, if you were ONLY doing a compression (set to maximum), would you do Jpeg, Jpeg 2000, or another?

 

[gordo]

So, if you were ONLY doing a compression (set to maximum), would you do Jpeg, Jpeg 2000, or another?

My understanding is that JPEG compression has a slight quality edge at low compression ratios (which is what you would be using), while JPEG2000 is better at high compression ratios. Basically, as you increase compression with JPEG the image starts to have blocky artifacts, but with JPEG2000 you don't see the blockiness so much but instead the image starts to blur.

I would stay with JPEG for compression at "high" or "maximum" quality with the least compression. You could always do a press test of images at different compression settings printed on a off-cut area of a press sheet (if that's available to you).

best gordo

 

[neleson]

Hi,


Nice discussion guys..But on balance i am totally agree with Gordo. Thanks for sharing Gordo.


epson t0715 multipack ink

 

[rich apollo]

So, if you were ONLY doing a compression (set to maximum), would you do Jpeg, Jpeg 2000, or another?

I steer folks away from JPEG because it is a lossy compression structure. I have had issues on press with JPEG artifacts. Gordo is right in his assessment that you probably won't have any issues, but 'once bitten, forever shy.'

JPEG2000 is not universally supported by RIPs. You run the risk of the file failing in the RIPping process and having to be edited. JPEG2000 is touted as being lossless, but Gordo's post contradicts that.

ZIP and LZW don't boast as great a compression ratio, or space savings; but they are lossless, safe, and universally supported (to my knowledge).

I resisted downsampling until recently. 300 ppi is a high enough resolution for almost any application. If 300 makes you nervous, try 400 ppi. Images that have been placed and reduced can easily end up exceeding 1000 ppi, and you will never see a benefit from that. I'm not sure that final rasterization on the RIP will be that high.

 

[Glenn McDowall]

I ran some tests once that downsampled high ppi images in pdfs to 300ppi. The one's that were jpegged were worse than the no compression. Re-applying jpeg compression to the 300ppi made them worse again. It also made a big difference which Application did the downsampling (quark pitstop and Acrobat's optimizer). The best way of getting these pictures into a pdf at 300ppi was always to let an Adobe application do it at source and not resample ppi after the jpeg had been applied first time round.
What I'm trying to say to the original poster is that if you are using the jpeg button out of InDesign then it may be best to downsample at this stage as well.

 

[kaiserwilhelm]

I will be doing it out of Acrobat. My PDF INTO my rip will still be huge. My pdf out of my rip will be huge. The pdf that goes to my customer., data processing (as a background image), and my new digital web press will be this jpeg.
Thus - convention press will have a perfect pdf. Customer a shrank. DP a shrank to show layout. All of those are fine. It is the digital web press that has me a bit concerned.

 

[Joe]

I prefer ZIP compression. It is a lossless compression so quality does not change. File size will be a little larger than jpeg compression but not by enough to be a concern for me. I also downsample anything over 450 dpi to 300 dpi and have never had an issue with that.

The latest Prinergy update supposedly fixed the JPEG2000 incompatibility and I've been using it for a few months now and haven't ran across any JPEG2000 problems. I used to get quite a few before the latest update.