Uncomp gives me 35mb file

[wparsonsgisnet]

Advertisement (gone after registration)

I saw a thread here, that I cannot find, talking about the size of the uncompressed dng. My file is 35mb.

 

I thot this was raised and someone indicated a setting that made it half that size -- without compression.

 

Am I imagining this?

[adan]

That's the normal size as created in the camera (uncompressed). 18 mpixels x 2 bytes (16 bits) per pixel x .98* = 35.28 Mbytes

 

(*always remembering that "megapixel" means an even 1,000,000 pixels, but a megabyte is 1,024,000 bytes, thanks to the wonders of binary math, so each megapixel equals about 0.98 megabyte in file size).

 

If you turn on the (lossy) in-camera compression for DNGs you get half the size.

 

After the fact, at the computer, you can save an open uncompressed DNG as a new DNG, at least in Adobe's raw developers, with lossless compression, and reduce the file size.

[35mmSummicron]

on a mac my M9 test files show up as

 

Uncompressed : 36.4 MB on disk (36,433,920 bytes)

 

Compressed : 18.3 MB on disk (18,324,480 bytes)

[chris_tribble]

This is the files size out of the M9 uncompressed. If you're concerned about storage / processing one option is to IMPORT AS DNG when you bring images off the card into lightroom. This drops the files size to the low 20s and, in my experience, speeds up the way that LR handles the images.

 

In practical work though (and this was supported by work done by others on the forum) there are few if any benefits from using uncompressed images above 160. My practice now is to have presets for 400 defaulting to compressed, reserving 160 as my UNcompressed setting. Works for me...

[vikasmg]

In practical work though (and this was supported by work done by others on the forum) there are few if any benefits from using uncompressed images above 160. My practice now is to have presets for 400 defaulting to compressed, reserving 160 as my UNcompressed setting. Works for me...

 

Chris, where you do use uncompressed files, do you see a signficant difference in image quality. What about for A2 and A1 size prints if you have tried them? I have not tried this yet, though as an architect I do have access to people with large format printers. So I probably will try a large print once I get my smudged sensor problem solved (but that's in another thread)

 

- Vikas

[chris_tribble]

I've not printed M9 images up to A1, but I have done so with M8 (up-resing in CR4) and had no problem with artifacts or smudging. Looking at the M9 images at 100% I can't see a difference between compressed and uncompressed (though I've not done any scientific tests...

 

I've not got anything that's really comparable, but as an idea, the first shot (135 Apo-Telyt 3.4 - who says you can't focus this lens!) was uncompressed. The second shot (50 lux asph) was compressed. In prints, I don't think you'd see a difference.

[wattsy]

Advertisement (gone after registration)

Late time of the year for lambs.

[chris_tribble]

Late time of the year for lambs.

 

Pyrennees in late September... and you're right - hadn't thought about that...

[35mmSummicron]

I LOVE how the M9 finally has the option to give FULL 14bit file output, but I also wish the M9 was fast enough to accommodate lossless data compression (clearly it is not).

 

at roughly 37mb per uncompressed raw, it is almost the same size as a Phase One P30+ IIQL lossless data compressed raw file (read 18mp vs. 30mp)

 

the advantage of lossless data compression is obviously to help conserve hard drive/SD card space, but also it makes less info to write presumably making it faster to write to card per shot........

[ho_co]

... where you do use uncompressed files, do you see a signficant difference in image quality....

Vikas, I don't mean to intrude, but if you missed it, there was a technical comparison thread on the topic of 'compressed vs uncompressed' M9 files earlier, where the person doing the tests demonstrated to the satisfaction of most of the readers (I would guess) that the advantage of the uncompressed version would be minimal, but if visible would show up only at ISO 160. (I didn't understand a word of it, but the graphs looked good. )

 

Hope that helps, though I know that citing mathematical comparisons isn't the same as comparing prints.