Go Back   Steve's Digicams Forums > Digital Cameras (Point and Shoot) > Panasonic / Leica

Reply
 
Thread Tools Search this Thread
Old Oct 26, 2005, 11:06 AM   #11
Member
 
AevnsGrandpa1's Avatar
 
Join Date: Dec 2004
Posts: 83
Default

After reading through the post it is an interesting question but I would like to point out a few things.

1. To get an accurace comparision between what you are calling the 2 and 4 megapixel mode, you would need to take pictures of the same thing, crop both to show the same thing and not introduct ANY modification to either picture.

How is it a fair comparison to manipulate one picture in Photoshop and not the other and then say the one without enhancments doesn't look as good?

2. If I am not mistaken, you terminology is incorrect when speaking about "2" and "4" megapixel mode. No where in the documentation this is said. The camera is a 4 megapixel camera. (I have it also), the different RESOLUTION choises are just that. The picture is still taken with the entire 4 megapixel CCD imager and then through the internal processing circuitry resized to the lesser resolution. This is no different, I think, than doing it in Photoshop by resizing the image.

I am not saying there is or is not a difference in the image quality when using the different resolution modes, but I would like to see you post as I described above, seeing 2 pictures with no manipulation.

If there is anything I was wrong in stating, please forgive.

Jeff
AevnsGrandpa1 is offline   Reply With Quote
Old Oct 26, 2005, 1:05 PM   #12
Junior Member
 
rubinsky's Avatar
 
Join Date: Apr 2005
Posts: 14
Default

Thanks for the reply Jeff. I stand corrected. What i was calling "megapixels",thecamera's settings menurefers to it as "Picture size".

Here are the two modes i waswriting about:

>Picture Size: 2304x1728 = 3,981,312 pixels (close to 4 megapixels)

>Picture Size:1600x1200= 1,920,000 pixels (close to2 megapixels)

Im well aware that the the camera will resample the CCD's maximum resolution to obtaina smallerimage. And I also know that if i shoot a 2 megapixel picture (1600x1200) there isnt a single thing in photoshop that could add 2x the level of detail needed to produce a 4megapixel image. Sure i can resample and interpolate from 2MP to 4MP but it wont make a "real" 4 megapixel picture, detail wise. Now heres the thing, I am getting the same level of detail when i compare the 2 modes, exept that in the highest resolution all i get is more noise.

I was hoping someone would metion it here but so far no one has. Ok heres what I was aiming for. I dont think this camera really puts out 4 megapixels, detail wise. It appears to be resampling and interpolating from a smaller resolution to higher one, because frankly, I dont see any more detail in "Picture Size = 2304x1728" (4MP) then i do in the smaller resolution of "Picture Size = 1600x1200" ( 2 MP) of my fz15.Not to mention that the small pattern of jagged lines appearing on very picture i take in 2304x1728 mode are a sure sine of poorly done interpolation the camera's internal circuitry, if it is indeed upconverting the resolution. Im not really saying or claimingthat is has a 2MP CCD. Im sure that it has a 4MP, but when it is possible to get a1600x1200 to have the same level of detail and less noise then the higher 2304x1728 mode, then something is wrong no matter how you put. Simply put, It should not be in anyway possible to make a 2 megapixel picture look as good as a4 megapixel picture, unless of course the 4 megapixelpicturestarted out asa 2 megapixel picture all along:G. Even then, it will still lack the detail of a "true" 4 megapixel picture.

This begs the next question. What is the true resolution of my fz15? It's certainly not 4MP but far less, maybe even half of what is claimed. Now i'm tempted to find a way to measure the true resolution of my fz15but given the factthat im currentlyliving in China, my resourses are prettylimited.

Anyway try this website it may help shed some light on what im talking about: http://megamyth.homestead.com/imageres.html

rubinsky is offline   Reply With Quote
Old Oct 26, 2005, 3:29 PM   #13
Senior Member
 
squirl033's Avatar
 
Join Date: Dec 2004
Posts: 6,720
Default

the "jaggies" you refer to appear in extreme closeup crops of almost any digital image, and are generally not visible in a print, unless you print only a very small area of the actual photo. the two images you posted appear almost identical, but as was pointed out earlier, you cannot compare an image that has been interpolated and processed in Photoshop with an untouched image and have a valid comparison. what happens if you process the 4MP image the same way?

i'm not an expert on the internal circuitry and how the FZ's resize from 4MP to 2MP when you change the image size mode, so i don't know, short of believing the pixel count in the image file itself, how you'd actually tell if the camera used a 4MP sendor or merely a 2MP upsampled. i tend to believe it's a 4MP sensor as advertised... there are too many risks of lawsuits, etc., over false advertising for a company like Panasonic to risk such a blatant misrepresentation.
squirl033 is offline   Reply With Quote
Old Oct 26, 2005, 5:29 PM   #14
Senior Member
 
Join Date: Nov 2004
Posts: 511
Default

Rubinsky, that was a rather dated article but not really the same ol' same ol'.
Plus it made sense of what you were saying to some degree. Whether I'm wasting card and computer space by using the higher value to record and store, I'd say I oft am. Some day I'll have to delete a bunch of garbage I have saved. No, I don't see an increase in jaggies when using a higher megapixel record.

What I have noticed is that that contrary to what I've been told longer time exposures result in far less noiseor its near elimination. Better focus also results in less noise in low light. Good exposures and focus works.

We're aware that sensors vary in their size, shape, spacing, arrangement and capability, whether they are capable of catching a single basic color or is even an affective pixel I'd say they count. The Panasonic sensors are pretty much as stated, not as what was it, Fuji?.
Tazzie is offline   Reply With Quote
Old Oct 26, 2005, 8:58 PM   #15
Senior Member
 
Join Date: Oct 2003
Posts: 141
Default

I am puzzled by these assertions. There is clearly more detail in the 4MP image than in the upsampled 2MP image. Look at the Chinese characters (or so I assume) on the spines of the books. Where you can see separate, distinct lines in the 4MP version, there are only undifferentiated blobs of black in the 2MP version.

Admittedly, the difference is not very dramatic, but that's precisely what the resolution difference amounts to.

The 2MP image may be more appealing to the eye, because it is, as stated, brighter. And the objects are slightly larger in this version. But for detail the native 4MP image is obviously superior.
Robb is offline   Reply With Quote
Old Oct 27, 2005, 12:14 PM   #16
Member
 
Join Date: Aug 2005
Posts: 38
Default

I toss my two cents in.

Its kind of hard to see the difference with objects that close to the camera.

For real test shots. Try doing a 4 & 2mp of a distant object with some detail. I found out that 2mp will run out of steam as the distant from the object increased.
Nightwing is offline   Reply With Quote
Old Oct 28, 2005, 5:42 AM   #17
Senior Member
 
Stoney79's Avatar
 
Join Date: Jan 2005
Posts: 314
Default

You show a 2 MP picture and a 4 MP picture, both resized to 1 MP. Clearly there's not that much difference. Please show us some crops of the original pictures, without any resizing.
Stoney79 is offline   Reply With Quote
Old Oct 28, 2005, 9:53 AM   #18
Junior Member
 
rubinsky's Avatar
 
Join Date: Apr 2005
Posts: 14
Default

Oh dear, I was hoping someone would attemp my experiment and then post their own results. On top of that I just realized that the pictures are slightly off axis so i have to re-do the experiment. But anyway, the results should be almost the same.


I would like to knowif anyone here knows why theynever implemented a raw picture format (excluding the fz30). It is quite easy to implement into the firmware and all it is, is just a raw dump of theCCD directly from the cameras catche and thenwritten directly to the sd memory with some minor encoding:? . I mean, the CCD is,electrically speaking,nothing more then a very large array of capacitive banks, and when a picture is taken,each bank collects photon energy which is measured for its independent voltage and assigned is a valueby an analog to digital converter which then quickly creates a raw-digital map of the picture that is then transfered to the onboard catche for further processing.

Could it be that they don't really want us to discover something fishy in the raw format, likesay for instance, a 2 or 3 megapixel raw dump!...oops did i say 2 or 3:G. Right now I can't really say 4:O. Oh well it's just unproven speculation of mine--hehe:G. Ok .Ok. I'm sorry I said that,I was out of line:?.


I am now working on a way to measure the actual pixel-count on the CCD of my fz15, but it will take me a few days to develop. Since im in china I will have to work with what I have, but at least I have a lot of free time, for the moment-- hehe. No internet shopping here though. I dont think the measure process will be 100% accurate but it would be interesting to see if anyone can reproduce my results.

Hi Stoney79,

If you read what I wrote at the beggining of this post you should have seen that the purpose of this experiment is to make an observation of wether a 2MP can be made to have practically the same level of detail as a 4MP picture, taken in two separate instaces with the same camera. In other words, I took 1600x1200 TIFF 2MP picture (not jpg) and I also took a 2304x1728 TIFF 4MP picture. I then up-sized the 2MP to 4 by using bicubic resampling and then sharpening it with "unharp mask" in photoshop in an effort to try to get it to look the same, or at least practically the same, as the untouched 4mp picture.

If indeed the two look similar then one could speculate two possibilities:

1. The camera is cheating by upsampling, which is highly unlikely.

2. The camera's 4mp CCD sensor is simply unable to deliver any more detail beyond 2 megapixelsdue tonoise or impurities.

To understand this process, one needs to understand resolution(size or MP) versus definition (quality and detail). And many should at least agree that no matter how large a picture is resized it will never gain additional definition or detail from the original master copy. It will just be a bloated copy of the original, and at best, it will loose some of it's definition in the process. It may help out your printer driver with interpolation and smoothing but thats about it.

So im comparing a bloated copy (enlarged and sharpened) of a 2mp picture against an original 4 megapixel piture. If anything the 4meg with an additional 2 million pixels should be able to easilly beat the 2mp by large margin. Just take any original 4mp picture and reduce it to 2mp. Then, make your best attemp to recreate the original using your choice of photo editing software. I can bet my life the 2mp will never have the quality and definition of the original 4 ever again.

Please understand that the two pictures were taken in separate instances, they are not derived out of one. No settings were altered during the 2 instances except for their sizes. Also please note that this post is an experiment for the purpose observation and debate. There are no conclusions in my observations as of yet. If you see any please let me know so I can make correction. If I do reach a conclusion, which is unlikely given my lack of resourses, I will put it up on another post for scrutiny.

In any case, i cannot provide the original file because of the large file size because, well, it's a huge TIFF (~11.4MB). About all I can do, is post the the 2 fullsize versions of the original because this post won't allow anything over 250kb for obvious reasons. So i will have to lower the quality. But at least if you get a full size- 4mp picture, you can superimpose it with the cropped 4mp on the top of the post by using photoshop and then at least you can decide for yourself if it is resized or not. Remember, that the 2mp on the top picture has been inflated to a 4 and then cropped to 1000x1000 so it makes no sense to compare it to the original 2mp ( you will end up comparing 1000x1000 against 500x500).

Please note: These are low quality jpegs in a zip file. Please do not--I repeat--do not use them to make any sort of quality comparisons. Sorry I blocked out the faces in the pictures but I do like my privacy. You can also view their exif for reference.

Thanks for all your feedback guys...


Heres the "full size" 4mp picture:







Attached Files
File Type: zip 4mp low quality.zip (137.8 KB, 89 views)
rubinsky is offline   Reply With Quote
Old Oct 28, 2005, 9:55 AM   #19
Junior Member
 
rubinsky's Avatar
 
Join Date: Apr 2005
Posts: 14
Default

Here's the 2mp low quality jpg:
Attached Files
File Type: zip 2mp low quality.zip (101.4 KB, 63 views)
rubinsky is offline   Reply With Quote
Old Oct 28, 2005, 2:14 PM   #20
Senior Member
 
squirl033's Avatar
 
Join Date: Dec 2004
Posts: 6,720
Default

Rubinsky wrote:

"If indeed the two look similar then one could speculate two possibilities:

1. The camera is cheating by upsampling, which is highly unlikely.

2. The camera's 4mp CCD sensor is simply unable to deliver any more detail beyond 2 megapixelsdue tonoise or impurities."

there's a third - and much more likely - possibility as well, which is that your eyes simply cannot resolve detail finely enough to see a perceptible difference between a 4MP unprocessed image and a 2MP file which has been upsampled to 4MPand sharpened.most people cannot see details (pixels)until the image resolution drops below about 150-170 ppi. if you take a 1600x1200 image and upsample it to 2304x1720, you've far exceeded the threshold of human perception (at least until the image is viewed at a size ofabout 11x16"or larger, at which point even the original 4MP file will start to show pixelization), and the only noticeable difference will be a slight "softness" in the 2MP image, due to the addition of interpolated pixels which may not have the correct color or brightness relative to the original. the end result is that an upsized and sharpened 2MP image may very well look so much like an unprocessed4MP image that you may not be able to see the difference. that is more a testament to an excellent resampling algorithm than it is to a flaw in the camera's sensor.

i have an 11x14 print on my wall, which i made from a 3.2MP image i'd cropped from the original 5MP file from my FZ20, then resized back up to 5.5MP and sharpened, just as you did. it's ever so slightly softer than the original 5MP photo in some places, but unless you looked closely - VERY closely - at both images, you would never know it wasn't printed from the original 5MP file. certainly from any normal viewing distance (24-36" or more), there is no perceptible difference detail or clarity.
squirl033 is offline   Reply With Quote
 
Reply


Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off



All times are GMT -5. The time now is 9:47 PM.