Jump to content

RAW image format foolishness


Recommended Posts

Why are digicams still use proprietary image compression and raw image

formats? The JPEG-2000 standard has been out for over 5 years now and

is a standard (and superior) image compression that is lossless. It

would seem that having a standard so that all the digital

infrastructure could work together (software, printers, etc.) without

having to have proprietary software and plugins would be a win for all

concerned. With proprietary formats, owning more than one brand of

digital camera seems like it is an unbelievable disaster to manage the

environment.

Link to comment
Share on other sites

Not all cameras and sensors are equal; even when they use the same RGGB pattern, the spacing may be different and there are lots of other variables.

 

From what I understand, Adobe "Digital Negative" allows for that and seems like a good idea to use as a standard. But raw is not the same as lossless compression, so it is different from what JPEG-2000 offers. JPEG-2000 is just a better JPEG; the camera would still need to process the raw and then turn it into JPEG-2000, losing some of the advantages of raw.

 

That said, that cameras still use good-old JPEG is pathetic, JPEG-2000 would be a big improvement, just not a raw replacement.

Link to comment
Share on other sites

What bas said. Comparing JPEG-2000 and RAW is silly, but the problem is a LOT of sectors

need to gain JPEG-2000 support. It would be nice if they hurried it up. I'd be happy with a

RAW->JPEG-2000 option in Adobe Camera RAW in a world where I could then do any number

of things with the JP2 file. But that isn't this world yet.

Link to comment
Share on other sites

That depends on what you mean by "truely raw". As best as I have been able to tell no

DSLR on the market today creates files of truely 'raw' data. You cite one example --

the compressd NEF format used nthe D70 cameras. here is anopth: Canon engineers a

mndatory sharpening routine routine into how all of their DSLR cameras process the raw

data before it is stored.

Link to comment
Share on other sites

In the mid 1980s Japan engineered what is now known as HDTV - a superior video format to what we are being sold in the US as HDTV. Today Japan is developing their next level of HDTV.

 

With multiple vendors on multiple platforms, politics plays a bigger role rather than who gets what level of technology. One vendor can create a standard and others in turn create their own standard. Oh and look, we have a dozen digital camera manufacturers each with their own 'raw' standard and I haven't seen one implement jpg2000 yet. It doesn't mean they haven't, just that I haven't seen it yet. There could be licensing fee's that Nikon, Canon, Sony, etc have to pay to resell JPG-2000.

 

I suspect it'll be 5 more years before we see something to compare/contrast against existing jpg compression. Your only other option is RAW. Take it or leave it.

 

Look at it this way, by the time we do standardize on HDTV, the Japanese and Europeans will be so far more advanced than the US that we'll be considered a 3rd world country with a HUGE deficit.

 

(yeah, thanks GW)

Link to comment
Share on other sites

As previously noted, the RAW format has little to do with compression loss. RAW allows full control of the image reconstruction after the shot. Some of the key parameters such as white balance, exposure compensation, sharpness, can be easily tweaked. There's a <a href="http://www.photo.net/learn/raw/" target="right">good article on this question</a> right here on photo.net.
Link to comment
Share on other sites

JPEG 2000 doesn't offer huge advantages and requires having powerful enough chips and new software to encode, so costs go up. Next generation DVDs are divided into two camps and there are something like four completely different 2G cellphone standards in the US, so I don't think the situation in photography will improve. And Ellis is right about "true raw".

 

But hey, IIRC Hasselblad's new bodies use DNG, so not all is lost.

Link to comment
Share on other sites

.

 

And while we're at it, why don't all cars use the same tail light lenses so I can get cheap replacement parts ... and how come toasters don't all use the same ... you know where this is going!

 

There are different "RAW" outputs (and RAW converter programs) because there are different features and benefits perceived by the manufacturers AND the consumers - and there are NEW features and benefits being developed every day.

 

I cannot imagine an amateur wanting (to pay for) the exact same features and benefits from their digital camera as a professional. Nor can I imagine a professional "settling" for the lowest common denominator features and benefits from our gear as an amateur!

 

So long as there's a reason for inovation and COMPETITION, we'll have choices - and I agree, making choices can be a pain sometimes, eh?

 

Unless someone wants someone else to make their decisions for them - as in Saudia Arabia, China, et cetera ... we all know where this is going!

 

Yes, it's hard to figure out the differences, and if the differences matter.

 

Alternatively, if RAW is too headdy a subject to master for one's own criteria, why not just "settle" on an existing standard: JPG/JFIF 8 bit pre-sharpened, pre-optimized, and so on?

 

Otherwise, RAW is just like being in the ol' fashioned darkroom where YOU da boss, and YOU decide how to mix your chemicals and set your temperatures and agitate fast or slow or in a swirl and so on. Ahh, the digital world ain't that far off from yesteryear after all, is it?

 

And, Joseph, I have at least 4 different RAW converters because I perceive different benefits from each of their competitive features, and that's just RAW converters! I also have at least 4 image manipulation programs because I perceive different benefits from each of their competitive features, and 4 printers because I perceive different benefits from each of their competitive features and so on. NONE is "best" - all have something to offer me for a variety of different situations.

 

Standardization? HELL NO, WE WON'T GO! ;-)

 

Click!

 

Love and hugs,

 

Peter Blaise peterblaise@yahoo.com http://www.peterblaisephotogrpahy.com/

Link to comment
Share on other sites

Old JPEG, especially with 1:1 chroma subsampling, is "good enough"

for most people, especially now that digicams often produce images

(in good lighting, anyway) that don't need color-correction editing.

 

Joseph, are you using JPEG 2000 to archive images? If not why not?

I'm not. PNG is faster to write and currently seems more universal.

It's possible that JPEG 2000 will turn out to be a total bust.

Link to comment
Share on other sites

To answer Bill's question... I use transparencies to archive my images. I've only acquired a digicam in the last few months (Olympus C-8080), and so far, my impression is that digital infrastructure is a huge mess. I haven't even gotten to the question of what format to use for archiving a digital image because I haven't really intended to archive them. But once I start using the digicam, darned if I don't suddenly have images needed archiving. I still shoot 35mm film for macro and flower photography, and medium format for other things I care about.

<p>

It will be interesting to see how many digitally archived images are no longer readable in 25-30 years, digital encoding is just one variable, media and media format another (8 inch floppies anyone?)

Link to comment
Share on other sites

1) RAW files store linearly encoded sensor data with only 1 color at each sensor site (the Sigma SD-9 & SD-10 are exceptions). RAW files commonly store 12-bit data.

 

2) JPEG-2000 stores gamma encoded image data with info about 3 color at every sensor site. JPEG-2000 stores 16-bit data.

 

Hence, I get:

 

Canon CRW: 4,955 KB

 

JPEG-2000: 18,658 KB

 

Both are lossless, both hold all the data. The RAW file stores 12-bits of losslessly compressed data for every sensor site. The JPEG-2000 file stores 48-bits of losslessly compressed data for every sensor site. So even though JPEG-2000 may have superior lossless compression, it has to store 4 times as much data. Add in the heavy CPU usage it requires and it would suck in a camera. Think of it taking nearly 4 times as long to write the file to your CF card and having the poor little in camera chip take 10, 20, or even 30 times as long to compress it.

 

Now, I could be wrong here, but it does not strike me as a helpful thing for digital shooters.

 

just one opinion,

 

Sean

Link to comment
Share on other sites

In response to Sean, the calculation you are doing assumes 1 pixel per sensor site even for cameras that only have 1 color per sensor. If one wanted to use jpeg-2000 for this there would be a pre-processing transformation that would pixelize the data generated by the sensor first, and the jpeg-2000 transform would apply to the pixelized data (3 colors per pixel).

<p>

In my original message I didn't mean to imply that JPEG-2000 should be the standard format for losslessly compressed data, just that it's been around for 5 years and the camera industry is still using proprietary compressed image formats.

<p>

Even if the standardized format were created on a computer after upload it would be better than today. If one wants to send a losslessy encoded image to a print business, what is the preferred format? Tiff? Seems pretty huge.

<p>

JPEG-2000 is like a discrete fourier transform over a wavelet basis, instead of the sine/cosine function basis normally used in the fourier transform. Because the wavelets have compact support, it is algorithmically cheaper than an FFT (fast fourier transform), so DSP chips to do it should not be expensive once the demand volume is in place.

Link to comment
Share on other sites

.

 

Previous poster wrote: "... In my original message I didn't mean to imply that JPEG-2000 should be the standard format for losslessly compressed data, just that it's been around for 5 years and the camera industry is still using proprietary compressed image formats ..."

 

No, the camera manufacturer's are NOT using proprietaty COMPRESSION formats as much as they are using proprietary ENCODING formats.

 

As said, why store 3 colors at 16 bits when only 1 shade at 12 bits was recorded?

 

Jpeg2000 apparently COST something to employ, and Adobe charges NOTHING for their (now owvership) of TIFF (from Aldus - remember - the key to growth is ACQUISITIONS!).

 

Hey, I still have 100% access to 5 1/4 inch diskettes (single sided, even!), the ol' Stacker hard drive compression (goes REAL fast on newer computers - and I can choose boot drives in CMOS), and ZIP files from years ago.

 

I see NO problem with all the RAW formats in evolution today!

 

Click!

 

Love and hugs,

 

Peter Blaise peterblaise@yahoo.com http://www.peterblaisephotography.com/

Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now
×
×
  • Create New...