THAT....... when you increase the ISO in camera that the small shot counter thingy reduces how many shots you can hold on your memory card? SO not only do you have noisier images but you can't take as many of them.......

ill stick with 100 then ;)

Which camera do you use?? I have the Canon 30D & noticed that minor detail. The amount of shots changes depending on how much information is captured. When you use ISO 1600 it (the sensor) is bringing in more light (and noise) thus more image info. You can shoot outdoors at ISO 100 and get less shots than indoors at ISO 800. Best to just check the image size after its captured and guess how many shots you have left. If I don't make sense, just check my avatar :tongue1:

short_ribs was curious about ...

when you increase the ISO in camera that the small shot counter thingy
reduces how many shots you can hold on your memory card?

I had not noticed that effect -- and would not have expected it, either. 

In order to be sure about what is happening, you should take test shots
of a scene at various different ISO settings.  As you make those test shots,
nothing but the ISO should change: same scene, same lighting, same
camera position, ... .  Then repeat that sequence of tests with a different
scene, or two.  Take notes that indicate the image number and ISO value. 

After you copy the image files to your PC's HDD, see if there is a variation
in file size that relates to one of the ISO settings. 

That will tell you if the ISO setting actually influences the image file size. 

Check all of the camera's settings -- you may find that changing the ISO
also changed something else ... possibly something even more important. 

--
Martin

Radlafx has pretty much nailed it.  The size of the JPG image file is determined by the amount of detail in the image...that's why you will note that not all your images out of a particular cam are the same size.
Noise will increase the file size, as will increasing the in-cam sharpening settings.  Also remember that the # of shots left is just the cams guess....the actual # will still vary depending upon the detail contained within those remaining shots.

I guess that really, saying there is more detail is not quite accurate....the cam picks up the same amount of info which is determined by the total # of pixels....but what happens is that compression is much more efficient with less differentiation of pixels.  That is why a JPG image of a plain blue wall will be much much smaller in size than a JPG image of grass...given the same JPG compression ratio.

So, if your cam does not compress RAW files (some do)...the RAW files "should" all be the same size....

yeah I realize what's happening just thought it's funny how that guestimation meter thingy changes for it.... Oh and feel free to mention other interesting things you've come across with cameras or anything photo related.... Oh and for those who asked I use a Canon 300D

@Radlafx...Interesting...I'll have to give this a bit more thought then.....

I know the RAW files out of my D70 are all different sizes...I attributed that to the fact that those RAW files ARE compressed in cam.
My understanding was that the RAW files out of the 20 & 30D were UN-compressed...but, perhaps there is some proprietary "processing" going on with the Canon RAW files as well.

There is additional info stored with each image other than sensor data that makes up the image file itself...various tags and exif data...but it shouldn't be that much different.

Try a little test...take both your 6 and 11 Meg RAW files and convert them to 16bit TIFF's...do they come out to "roughly" the same size?

Attached Link: sensor size to file size calcs....

Some good points, image size can increase depending on the number of colours file size can vary a lot... To be honest in carrying extra cards around it is not much of a problem, it may be interesting to know how much it could effect the frame rate

It is ironic, isn't it, that the more noise you have in an image, the more data the file contains. I've noticed the difference many a time myself. Save a noisy image web size with no manipulations then run it through noise reduction software and save it again. That file should be noticabley smaller.

Attached Link: http://www.dpreview.com/learn/?/key=noise

it may be interesting to know how much it could effect the frame rate

With your exp Danob, you do surprise me.

A valid point though as more data should translate to longer write times, no matter how miniscule.

Shouldnt think it matters on the 5D Kort...3 fps!
Never bothered with that...always 1 shot on that body.
But, again, it depends on how faster lens one has at the end!

On my D70 it makes no difference whatsoever in the counter if I change the ISO from 200 to 1600.

Ok....It was late last night when I post previously....got some sleep...and thought some more...

Of course most RAW files are compressed!!!....the Canon crw, cr2...the Nikon nef, the Oly orf....etc, etc....they  ALL contain proprietary compression algorithms. Some are lossy, some are lossless.
So, their RAW file size is not indicative of the true amount of data captured by the sensor.

I thought though that Canon did have a feature to allow an UN-compressed file to be written to the card....perhaps some of the Canon guys can jump in here.  Some of the Nikons allow TIFF files to be written. The RAW format was developed to create a more efficient ( faster) way of writing data to the camera's card.

Regardless, point is...there is not "more" data with a noisy image....the amount of data captured is a function of the sensor.  The sensor captures the same amount of data regardless of scene (see my post in the sensor size/file size calc link above)...it is the compression algorithm (either JPG or RAW) that alters the file size.  "Similar" data sets allow for easier compression and more efficient file sizes.

@Kort...there really isn't more data...but the way the cam compresses that data into the RAW file determines the size...so, yes, a larger RAW file will take longer to write to the card.  If you wrote TIFF files to the card...the file sizes should all be "roughly" the same size.....although the write time would be painfully slow.  Again, do a test...convert your RAW files to TIFF...take a look at the bottom of your PS window.

Thanks for clearing that up! ;']

Just to add a bit more....each pixel of the sensor picks up data on each of 3 channels (R,G,B) and it also picks up 12 bits per channel.

The sensor does not differentiate between a noisy image...or a non-noisy image...the data from each pixel is sent to the computers processor.  Data is collected from all the cam's pixels....each pixel provides the same amount of data regardless of scene.

The sensor doesn't care whether that data is noise...or apple pie....the data gets collected.

If your sensor is NOT picking up this data...you have a problem.

Data is data, 1s and 0s. It is the camera's compression settings that decided how it is stored and how much space it will takeup. Non-compressed, all images would take up the same ammount of space, no matter how much or how little detail in the file. Compressed, it varies by manufacturer. Does that about sum it up?

Greetings to all,

This is becoming an interesting topic of discussion. 

I Googled for the search terms 'canon raw files compression' ... and
found that there are 710,000 web pages with those search terms!

I reviewed some of those pages and found these unequivocal but
unequivalent statements ...

Since there is no 'lossy' compression, the Raw file contains the full
data presented at the camera's sensor: there are no JPEG or other
compression artifacts.

[http://www.dlcphotography.net/RawFormatWorkflow/RawFormatWorkflow.htm]

A quick glance at a folder of 150 D30 images gives sizes ranging from
2,509Kb to 4,511Kb. The RAW format does include compression (heck
on the D30 it even includes a 2/3 size JPEG of the image as well!), so
different images will compress to greater or lesser amounts. I would
guess the extremes of file size range are in the region of between 2 and 5Mb.

[http://www.binbooks.com/books/photo/i/l/57646AFA31]

Lastly, JPEG compression used in cameras is a "lossy" format meaning
that some data is lost in compressing the file. Shooting RAW allows you
to bypass this compression and retain the maximum amount of data in
your images.

[http://www.mkwphotography.com/workflow.htm]

After seeing these seemingly contradictory opinions, I wouldn't dare state
that Canon Raw format either is or is not compressed ... unless I had a
highly detailed description of the Raw format directly from Canon.

--
Martin

p.s. Yes, I have the technical background to understand the actual format
that Canon uses.

@Kort...you got it!

Ok...now I'm off to have another slice of apple pie.....

Lol what would life be like without a slice of apple pie

@MGD;
It has to be compressed...do the math your self.
If it wasn't, a 6MP sensor would produce around 27MB of data.
To simplify the calcs in my link;
6,000,000 X 3 (channels) X 12 bits per channel = 216 million bits
Now remember, 1 byte = 8 bits...so 216M/8 = 27 million bytes.....although not exactly, but for sake of argument....27MB.

So you can see there has to be compression applied....
Whether that compression is truly "lossless"....well, that is something that only the camera manufacturer will be able to say...and most ain't sayin'....

Back to my pie....

Hmmm f4 lens are not that fast to make any differenece to the frame rate !! To get at the Raw file compression files of Canon Nikon cameras you need a util called DCRAW a free sourcecode file you can find here   http://www.cybercom.net/%7Edcoffin/dcraw/ Most of the major players also decrypt the metadata whatever the above link is an interesting read for those who want to contribute something worthwhile to the thread.. You may find it interesting  Thanks to those who have..

My wife makes a mean apple pie...have to see if she is planning one for tonight... Though I am a total noob when it comes to this stuff, I am guessing lossless compression looks at the image data and when there are patches of pixels that have the same color data it groups them together to conserve space, like patches of blue sky. It is just a theory, but seems to make sense -- at least until one of you good people shoot it down, hahahah.

TwoPynts indicated that he was curious about

lossless compression

Compression of a computer file reduces the size of the file.  There are
dozens of compression methods.  For each of these, there is also a
decompression method. 

The most important attribute of a compression method is 'loss'. 

Compression and loss do not always go hand in hand. 

1. A 'lossless compression' method does not loose any of the information
   that was present in the original file. 

e.g. the PKZIP program uses lossless compression methods.  GIF, PSD
and PNG are non-lossy compression formats. 

1. By contrast [BTW, would this use of the term 'contrast' create a puntax
   liability for me?], a 'lossy compression' method does not preserve all of the
   original information -- only just enough to fool you into thinking you got
   back what you started with.  Some of these lossy compression methods
   have (allow) varying degrees of 'loss' ... some implementations allow you
   to choose how much loss ... i.e. how much quality -- some implementations
   assume that we're just too dumb to know what to do and don't allow us a
   choice.  Some marketing types prefer to spread confusion instead of straight
   answers. 

e.g. JPEG and MP3 are always lossy methods.  How much loss?  Well, that depends ...

TwoPynts also guessed that

looks at the image data and when there are patches of pixels that
have the same

... yes, that is an aproximate description of compression methods in general. 

I had to laugh when TwoPynts said,

Though I am a total noob

and

at least until one of you good people shoot it down, hahahah.

... because he did a pretty good job of tricking us into explaining it.  LOL

--
Martin

Hey Martin, no peeking behind the curtain! :b_uncertain: :m_tongue2: STUFFIT also preserves all of the data intact, but that leads to the question (since you spoke mostly of lossy above) -- how does lossless compression work? TIFF files everywhere want to know exactly what the LZW is doing, hahahah.

Attached Link: Data compression [Wikipedia]

how does lossless compression work?

The Wikipedia article will get you started. 

Extra credit will be given for reading and explaining some of these additional
jucy topics: algorithmic complexity theory, information entropy, image
compression, speech compression, video compression, Huffman coding,
Adaptive Huffman coding, discrete cosine transform, fractal compression or
transform, wavelet compression, A-law and Mu-law Companders. 

--
Martin

hint: A-law and Mu-law Companders appear in telephony applications. 

@Kort....I'm sure I can find room for another slice of apple pie.
Tell your wife that there will be one more coming for dessert.....

Sorry, I ate it all myself! ;']

And when all the tech talk has your head twisting, take the X700 loaded with film of choice and hit the great outdoors! Something to do while your brain assimilates the information. No compression, noise, etc. to worry about, just focus, framing, and composition.