KS-2 DNG vs PEF

Some tests on compressing/decompressing HDD files...

http://directedge.us/content/to-compress-or-not-to-compress-part-ii

--

Interesting, but very much windows/office/general computing with generic compression that doesn't suit all things, especially raw data from a sensor.
I'm talking about my own vaguely remembered observations using cameras with relatively small buffers and relatively small SD cards (8, 16, 32 MB) and JPEGs TIFFs RAWs on an SSD (hard drive) being opened up in Photoshop or similar.
I suspect the results will be very different from the example above. But I'm happy to be proved wrong as I use DNGs anyway

Ha! Ha! I get it now..... Generally doesn't mean generally and disk doesn't mean disk

I can now go down to the pub and have a few pints

--

Tomay-to, tomah-to. I have no rationale for using PEFs other than I'm just comfortable with them as they have served me well since 2004 and my *istDS. DCU5 opens those RAWs perfectly well and will I suspect continue to do so into the future, along with those that I take now. If I want to make a DNG to, say, process with Camera Raw, then I can either convert it using the free Adobe converter, or convert it direct in DCU5.

So we are offered a choice of which format to use at source, which is great. Why is having the choice of both PEF and DNG formats available after the fact more restrictive than just having DNG?

Has anyone done a file size test taking the identical picture in PEF and DNG?
(I suppose I could but have never got around to it)

For those unfortunate not to have a camera which shoots DNG (Canon, Nikon etc), Scott Kelby in his books always recommends on import converting the cameras native RAW file to DNG. On of the advantages he says, results in a smaller file size.

No, never actually got that far! Yesterday, I formatted both cards in my K-3II and after that with RAW set to PEF, card 1 showed 600 shots remaining, and then after changing to DNG, it showed 599. Maybe I'll do a comparison today.

I spoke to a client of mine about it who is a pro sports photographer, when he was a hobbyist he used Pentax but is now a Nikon guy as he moved to FF.

His belief is that Camera manufacturers develop their own raw formats for faster write speeds which enable better fps in burst or continuous shooting. He suggests if you shoot a lot of burst and continuous as he does you will notice that you get less lag and marginally faster shooting. He reckons it makes next to no difference to the average user but often improves what they can publish on the spec sheet.

He says that most proprietary raw files are bigger than dng but that PEF in the latest Pentax is marginally smaller.

He suggests that there is no difference in image quality. And that he generally is happy to use PEF as he doesn't believe the proprietary formats will disappear as although they are not open source like DNG manufacturers happily share them with software manufacturers on a non disclosure basis without charge as it is in their interest's that the format is easily worked with.

But if you are worried and don't need the marginal improvement in in camera processing you won't lose out by using DNG.

The explanation makes perfect sense to me as I had wondered why manufacturers bother spending money developing their own formats if DNG is all that.

Thanks, John, sounds very plausible.

Shooting the same static subject, the PEF was 29.6MB, the DNG 30.2MB, so the latest PEF is indeed slightly smaller. What's also interesting is that up to the K-5, the Microsoft add-on Windows RAW codec pack happily displayed thumbnails in picture folders, but with the new PEF in the K-3II, it doesn't do that any more. Using the Adobe codec, the DNG does display as a thumbnail as it always has, but it's slow. It's not a real issue since I use Faststone (which displays both) instead of Windows Explorer to view the contents of folders.

So it looks like the PEF is not really compressed at all, just different. The DNG probably just contains more generic, less efficient code.