Low light noise of My E-500

[OlyFlyer]

This Japaneese site http://digitalcamera.impress.co.jp/06_02/auth/toku1/index_iso.htm is presenting a worthless test of several cameras, amongst which, one is the E-500. The test is worthless because so far nobody explained how the tests were done. The result for Oly at ISO 1600 is not very pleasing, due to all the colored dots in the white figure against the black background. So, I decided to do my own 'worthless' test of about the same settings.

My settings are as follows:

Light: 10W 12V halogen spotlight from about 0.5 metres from the object.
Object: White chess figure, 4cm heighth.
Background: Black cardboard paper.
ISO: 100, 200, 400, 800, 1600
NR on, RAW format used.

Lens 1: 14-45 mm at 45mm
Autofocus used

Lens 2: Zuiko OM 50mm f1.8 set to f16 + MF-1
Manual focus (obviously)
Aperture priority
Central weight measurment

I took 33 images, will post some tomorrow because now it is after 11pm and I am tired.

 

[eow]

This Japaneese site http://digitalcamera.impress.co.jp/06_02/auth/toku1/index_iso.htm is presenting a worthless test of several cameras, amongst which, one is the E-500. The test is worthless because so far nobody explained how the tests were done. The result for Oly at ISO 1600 is not very pleasing, due to all the colored dots in the white figure against the black background. So, I decided to do my own 'worthless' test of about the same settings.

My settings are as follows:

Light: 10W 12V halogen spotlight from about 0.5 metres from the object.
Object: White chess figure, 4cm heighth.
Background: Black cardboard paper.
ISO: 100, 200, 400, 800, 1600
NR on, RAW format used.

Lens 1: 14-45 mm at 45mm
Autofocus used

Lens 2: Zuiko 50mm f1.8 set to f16 + MF-1
Manual focus (obviously)
Aperture priority
Central weight measurment

I took 33 images, will post some tomorrow because now it is after 11pm and I am tired.

using only E-500 it wouldn't count
the best test would be when u had others brand to compare with....

 

[drektster]

strange, the picture taken by e300 at high iso is actually better than e500.

 

[Clockunder]

Olympus DSLR, with its 4/3 format which has a smaller image sensor, is generally known to have higher noise issues than Canon and Nikon DSLRs.

 

[ykkok]

Olympus DSLR, with its 4/3 format which has a smaller image sensor, is generally known to have higher noise issues than Canon and Nikon DSLRs.

That also depends on the number of pixel (photosite) on the sensor.

An 8MP sensor for APS-C size sensor has about the same sized pixel of 5MP 4/3 sensor.

APS-C is 75% of 35mm format, 4/3 is 50%. Looking at these figures, APS-C is not really that much bigger.

On top of that, S/N ration also depends on the sensor design. So does NR implementation.

 

[mpenza]

APS-C is 75% of 35mm format, 4/3 is 50%. Looking at these figures, APS-C is not really that much bigger.

You have to look at the area....APS-C sensor is around 40-44% the size of 35mm format and 4/3 is around 25%. APS-C is almost 60% larger than 4/3.

 

[stk]

IMHO, the strengths of olympus are their natural color rendition, quality of their budget zuiko lenses and ultrasonic anti-dust technology.. Due to the smaller size of the 4/3 ccd and squeezing 8mp onto it, higher ISO performance will not be as good as APS-C size competitors... ;p

U can't win at everything..

 

[OlyFlyer]

using only E-500 it wouldn't count
the best test would be when u had others brand to compare with....

It does indeed counts. I mean, I know, I can not say how E-500 is compared to other cameras. I am not really interested in comparing my camera with Nikon, Canon or not even other Oly cameras at the moment. I can not even say how E-500 is, just how MY camera is performing for me. What it says is: Am I happy with MY results? My intentions are not to buy an other camera and find out which is the best, I just want to learn my cameras limits and see if my image looks as this Jap site shows. I don't even belive that comparing a lot of different cameras would be valuable using only one type of image. I don't think there is a single soul on earth who buys a camera just to take pictures of white chess figures at low light against black background. One ususally selects camera based on many other facts/feelings than low light high ISO performance.

And since that is the only camera I have that is the only one I can test and interested in right now.

 

[OlyFlyer]

To Everybody who reads this,

I started this thread just to present you the findings I make with my E-500. I am not interested in why Oly is noisier than others, which brand is best at high ISO low light, if APS-C is better than 4/3, CMOS agianst traditional... and so on. That is already discussed to the end, nothing new to say. If that is going to be repeated I just remove my thread, because that is boring. A lot of CS ppl are talking like big time experts knowing everything about every camera but basing their 'facts' on single tests they don't understand. Some CS ppl sound like they own every camera in the world. I an not interested in reading these 'experts' comments. And I have to admit the sad fact, I only have one DSLR, that is my E-500. If I had more than one DSLR I would do the test with all of them, but that would still not be a scientific test. What I want to do is to see how my E-500 is compared to other film SLRs I had in my life, and how it compares to film in general. If ppl are interested, I will show what I find but I don't want to discuss how Sony Alpha or Nikon D200 is outperforming Oly E-500 at low light. Comments from ppl who actually are owners of other brands are welcome, if these are based on own experience and if they are not braging about how bad Oly is and how good their camera brand is.

Doing tests like these are technically and scientifically worthless because of many facts. For example, only one body is used. Temperature is raising in the room since I work with closed doors and windows to exclude exterior light because in Sweden it is never really dark during the summer season. In fact, the room temperatore was at start 21degrees and it got to 29degrees when I was ready, which definitly effects image quality. Also due to the fact that I used a halogen light near the camera, I think the temperature in my cam was much higher. Actually, I found that the 10W light is just too bright, so I will try to reduce that by increasing the light - subject distance or using white LEDs instead of halogen the next time to get real long time. Now the longenst time was 1.3s using f22 with ZD lens and f16 with OM.

 

[ykkok]

You have to look at the area....APS-C sensor is around 40-44% the size of 35mm format and 4/3 is around 25%. APS-C is almost 60% larger than 4/3.

You are right, my mistake. By area it is.

Olyflyer, we are not comparing on this anymore but just that there are many misconception here that people mistakenly think that 4/3 sensor is dinky compared to APS-C.

The size of the sensor does matter, but the density matters too.

 

[OlyFlyer]

You are right, my mistake. By area it is.

Olyflyer, we are not comparing on this anymore but just that there are many misconception here that people mistakenly think that 4/3 sensor is dinky compared to APS-C.

The size of the sensor does matter, but the density matters too.

Don't misunderstand me, I am interested in reading facts. Just that I am not iterested in hearing how crappy Olympus cameras are, and how others are better under low light. Personally, I know the differences technically between different CCDs and technologies. I have also read a great deal about noise before, and included links to other sites on other threads where the subject was discussed before, to be read by those who have a higher interest for knowledge. My only interest here is to see how my camera performs and to share that with everybody who wants to see my results. I don't want to convince anybody about E-500 is good and I don't want to be convinced E-500 is bad. It is that part I find boring. Reading about facts like sensor size differences, S/N, pixel size and resolution is nothing that I have anything against.

BTW, I think the Oly sensor is 'dinky' in the sense that it is smaller than APS-C, looking at their area only. But I wonder how many have today a screen that has the same format as APS or how many prints in APS format. If the full size of the screen or paper is used, I think you have to resize an APS so much that you actually get a considerably smaller image size with large border at top and bottom, or you loose a large number of valuable pixels left and right. I have not calculated that, but comparing simply the sizes can be done differently.

 

[Clockunder]

That also depends on the number of pixel (photosite) on the sensor.

An 8MP sensor for APS-C size sensor has about the same sized pixel of 5MP 4/3 sensor.

APS-C is 75% of 35mm format, 4/3 is 50%. Looking at these figures, APS-C is not really that much bigger.

On top of that, S/N ration also depends on the sensor design. So does NR implementation.

This means that the pixel density of the 8MP Olympus E-500 image sensor is equivalent to a 12MP APS-C sized image sensor and is therefore denser than a 6MP Nikon D70 or 8MP Canon 350D etc.

All else equal, higher density means higher susceptibility to noise.

 

[OlyFlyer]

Here are the first images. To give you some dimensions, my figure has a total height of 4cm. The top circular part has a diameter of 9mm, the horizontal part in the middle of the crown is 8x1.5mm in size. The foot has 15mm diameter. Original color is creamy white. I did not apply any corrections in computer, did not try to improve image, just converted to JPEG and cropped. NR was enabled in camera.

#1 is the full size chess figure. Unfortunatly, CS resizes the picture and does a new compression, which is not good for the image.

#2 is the top part of #1
This was not resized and probably not re-compressed by CS because it has the same file size as my original. This is a 100% cropped piece of the original image.

So far all ISO 100 but I have to say, I am happy with the images produced. Color is also very nice, which is amazing considering the black background and the creamy white figure. Auto WB was used. Also the AF was working without any problem considering the low light, which can be seen in the details. Actually I used f22 for these images not to get larger DOF but to get longer time. This way I could get 1.3 sec for the above images.

More picture tomorrow, about the same time, around 2300 hour (Swedish local time).

Playing with Olympus Studio using my 1600 ISO image definitly tells me I could never have taken better pictures with my (or any?) film camera ever under these really bad conditions. So, by now I know that if E-500 beats my previous cameras, than I am happy with E-500. We will see what happens in the long run.

 

[ykkok]

Olyflyer,

Try max f16, not f22. It'll be sharper.
As for DOF wise, due to the 4/3 shorter equivalent focal length, f16 is almost f32 in 35mm format.


Clockunder, yes you are right. One way to maintain good s/n is to increase the sensor size if the #of pixel increases, like 16MP on 35mm format.

 

[OlyFlyer]

Olyflyer,

Try max f16, not f22. It'll be sharper.

How come? Shouldn't it be sharpest with smallest aperture? Or is this something ZD special? :dunno: I will try next time. Anyway, I used f22 to get long time, so with f16 I definitly have to move back the light to get about the same conditions. I was quite happy with my image, considering the conditions and you say the image could be made even better. Thats good news

 

[Hacker]

How come? Shouldn't it be sharpest with smallest aperture? Or is this something ZD special? :dunno: I will try next time. Anyway, I used f22 to get long time, so with f16 I definitly have to move back the light to get about the same conditions. I was quite happy with my image, considering the conditions and you say the image could be made even better. Thats good news

The smallest aperture has nothing to do with a picture being sharp at a given focusing distance. For example in the ZD50mm case, the sharpest or optimal setting is aperture f4 for the center and f5.6 for the border. F22 is the worst possible setting for both center and border.

http://www.photozone.de/8Reviews/lenses/olympus_50_2/index.htm

 

[mpenza]

It's been said that diffraction effect will reduce sharpness for small aperture. Not too sure how it happens though.

 

[AQVA]

its like your eyes... u squint your eyes slightly and get a better and sharper vision, but once u squint further your vision will be blured.

 

[OlyFlyer]

The smallest aperture has nothing to do with a picture being sharp at a given focusing distance. For example in the ZD50mm case, the sharpest or optimal setting is aperture f4 for the center and f5.6 for the border. F22 is the worst possible setting for both center and border.

OK, but that is still only for 50mm, maybe only for that individual. Or are these values valid generally for all ZD50mm? Can we assume Oly quality is even? However it is, it is still very interesting to see his values.

For the lens I used, his verdict is very good, he says "The lens showed a very good performance throughout the range and and all aperture settings. Seems as if Olympus managed to squeeze quite a bit of quality from the bottom end kit lens - amazing for a lens of this price class." Unfortunatly he presents only f5.6 and f8 for the sharpness, no warning about f22 being less sharp than f16. Anyway, I will redo my tests using f16.

http://www.photozone.de/8Reviews/lenses/olympus_50_2/index.htm

Thanks! Very interesting site, a lot of useful facts to read. Link ended up in my Explorer Favorites. :thumbsup:

 

[OlyFlyer]

It's been said that diffraction effect will reduce sharpness for small aperture. Not too sure how it happens though.

I think it is the fact that light rays bend around corners more distinctly when small aperture is used. But then f8 should be even better than f16.

Actually, f8 would give me problems with DOF. According to my calculator, 45mm lens, 0.5m to object, f16 gives 5.3cm DOF. Using f8 would give me only 2.7cm DOF, so I guess I just stick to f16 as ykkok proposes.