The GetDPI Photography Forum

Great to see you here. Join our insightful photographic forum today and start tapping into a huge wealth of photographic knowledge. Completing our simple registration process will allow you to gain access to exclusive content, add your own topics and posts, share your work and connect with other members through your own private inbox! And don’t forget to say hi!

GF1 dynamic range

J

jerryk

Guest
Hi,

Has anyone else noticed that the GF1 seems quite limited in dynamic range? I have been quite surprised by the number of shots that get blown out with this camera.

Jerry
 
C

Chuck A

Guest
Jerry,

It happens very often that general statements like this are made without any support. There are so many variables involved here that you should include photos and even do some exposure bracketing to show the problem. While I agree that the GF1 has a limited dynamic range it is certainly not horrible and with a little care and shooting RAW it is actually pretty good. I don't know you or your photography but examples might help.
 
N

noirist

Guest
Well said, Chuck. For what it's worth, Barry Green at dvxuser tested the dynamic range of the GH1 and the 7D and found the GH1 was slightly superior.
 

Rawfa

Active member
Well said, Chuck. For what it's worth, Barry Green at dvxuser tested the dynamic range of the GH1 and the 7D and found the GH1 was slightly superior.
I have over 2000 posts at dvxuser and even though Barry has lots of experience in the video world I would not trust his reviews/opinions completely. The website is sponsored by Panasonic and Barry's dvx dvd is sold along with panasonic cameras. This has been a point of debate in many MANY other video forums and blogs.
 

Rawfa

Active member
What's I've found out shooting with micro 4/3 is that you REALLY have to shoot RAW to get the best out of this system.
 

pellicle

New member
Hi

its an interesting point ... while there is great usefulness in understanding your tools, but just how do they arrive at the measurements they publish? Are they using JPG or RAW and what is their sample methods? I haven't found what the methods of DPReview are ... clearly they are influential in shaping peoples opinions.

For a start I put this post up the other day, although noone seemed to find it worth commenting on.

That data (from dpreview) seems to show that the GF-1 has a wider range than the G1. Other evidence suggests that they are different sensors.

My own personal experiments in comparing my G1 to my 10D indicated to me that they were really close, but I didn't make 'measurements'

fundamentally I think nothing is perfect and people (including me) have been quite satisfied with the tools we've had to this point. Many people seem to fluff up their black and white film exposure (which is hard to do) enough and still get good prints

so why not just pick the camera based on the important cirteria? Like ... does it have the look and feel you're after? Does it have the flexibility you look for (say in interchangable lenses)

if so, go grab it
:)
 

pellicle

New member
What's I've found out shooting with micro 4/3 is that you REALLY have to shoot RAW to get the best out of this system.
well I agree with that on any system .. :) but the Olympus seems to do pretty well out of the box
 

ggibson

Well-known member
Agreed on shooting RAW. I've only had a couple of shots with significant blow-out. Compared to my old V1 which only shot jpg, the dynamic range is great. Here's one that I took facing directly into the sun:



This is post-processed in Lightroom 2.5, where I pulled up the shadows and brought as much sky back as possible. Not bad, I think, considering the limits of the scene. Here's another one I shot that day with bright white boats reflecting the sun at noon:

 
J

jerryk

Guest
GGibson,

Is the marina Emeryville? I work nearby the marina in Emeryville.

Jerry
 

Rawfa

Active member
Take a look at any picture I've posted before I've started using the E-P1 and you'll find out. While I've found out with other systems that shooting RAW would improve your possibilities, I've never seen a night and day improvement until I started using the E-P1 (for the first time I HAD to shoot RAW to get the results I've wanted).
 

ggibson

Well-known member
GGibson,

Is the marina Emeryville? I work nearby the marina in Emeryville.

Jerry
Nope, but pretty close--it's right behind AT&T Park where I took the other picture. I work nearby and went for a walk-around at lunch :thumbup:
 

Godfrey

Well-known member
Has anyone else noticed that the GF1 seems quite limited in dynamic range? I have been quite surprised by the number of shots that get blown out with this camera.
I've heard this since the G1 model released last November. Several months ago I decided to do some DR and meter calibration testing, comparing my E-1, L1, K10D, G1 ... and a week later had a D200 to put through the same test as well.

What I found was that the G1's ISO sensitivity setting is 0.3-0.5EV pessimistic compared to the E-1/L1/K10D and the D200's sensitivity about 0.5EV optimistic. What this means is that if you set ISO 100 on all five cameras, the actual sensitivity measured by a calibrated reference incident meter and gray card target indicates that the G1 sensor is actually at ISO 140-150, the Nikon is actually at about ISO 50-64, and the other three are with 5% of 100.

The next thing I noticed was that the metering characteristics of the cameras varied quite a bit too. The E-1 was most protective of highlight saturation (it would suggest about -0.7EV compared to a nominal reference reading), the L1 and K10D were closer to accurate at -0.3, the D200 showed nominal at 0, and the G1 tended to be right on the line or a little over at +0.3 EV.

These mapped pretty closely to my picture taking experience where I find I normally have the E-1 set to +0.7EV compensation, the K10D and L1 set to +0.3EV, the D200 at 0EV, and the G1 at -0.3 to achieve the same average mid-gray point with the RAW processor at the defaults.

Once I dialed in these corrections and ran the DR test, all five cameras showed remarkably consistent DR performance. Absolute numbers are not really very interesting, but relative performance is useful since they were all tested in the same circumstances using the same methodology .. DR decreases as ISO goes up, so the numbers are presented in stops as {lowest ISO} to {highest ISO}:

K10D: 10.5 to 9
L1, E-1: 10 to 8.5
G1: 10 to 7.5
D200: 9.5 to 8

So ... The G1's DR is fairly comparable to similar format DSLRs, with a little hit at the highest ISO settings. However, the combination of the most pessimistic ISO sensitivity setting and metering calibration that pushes the saturation exposure limit will artificially restrict DR if you don't adjust take the hot sensitivity and metering calibration into account.

The GF1 is the same sensor, same metering, and almost the same data system ... I'm sure the results for it would be very similar to the G1.
 
C

Chuck A

Guest
I've heard this since the G1 model released last November. Several months ago I decided to do some DR and meter calibration testing, comparing my E-1, L1, K10D, G1 ... and a week later had a D200 to put through the same test as well.

What I found was that the G1's ISO sensitivity setting is 0.3-0.5EV pessimistic compared to the E-1/L1/K10D and the D200's sensitivity about 0.5EV optimistic. What this means is that if you set ISO 100 on all five cameras, the actual sensitivity measured by a calibrated reference incident meter and gray card target indicates that the G1 sensor is actually at ISO 140-150, the Nikon is actually at about ISO 50-64, and the other three are with 5% of 100.

The next thing I noticed was that the metering characteristics of the cameras varied quite a bit too. The E-1 was most protective of highlight saturation (it would suggest about -0.7EV compared to a nominal reference reading), the L1 and K10D were closer to accurate at -0.3, the D200 showed nominal at 0, and the G1 tended to be right on the line or a little over at +0.3 EV.

These mapped pretty closely to my picture taking experience where I find I normally have the E-1 set to +0.7EV compensation, the K10D and L1 set to +0.3EV, the D200 at 0EV, and the G1 at -0.3 to achieve the same average mid-gray point with the RAW processor at the defaults.

Once I dialed in these corrections and ran the DR test, all five cameras showed remarkably consistent DR performance. Absolute numbers are not really very interesting, but relative performance is useful since they were all tested in the same circumstances using the same methodology .. DR decreases as ISO goes up, so the numbers are presented in stops as {lowest ISO} to {highest ISO}:

K10D: 10.5 to 9
L1, E-1: 10 to 8.5
G1: 10 to 7.5
D200: 9.5 to 8

So ... The G1's DR is fairly comparable to similar format DSLRs, with a little hit at the highest ISO settings. However, the combination of the most pessimistic ISO sensitivity setting and metering calibration that pushes the saturation exposure limit will artificially restrict DR if you don't adjust take the hot sensitivity and metering calibration into account.

The GF1 is the same sensor, same metering, and almost the same data system ... I'm sure the results for it would be very similar to the G1.
Great post! You do the tests that I can't seem to find the time to do. Funny, I have had a GF1 and G1 for about a month and have noticed without testing what your tests show. I routinely dial down to -.3 to get the exposure that I want.

So, as far as my request from the first post. Making a statement that this or that camera is inferior in some way without any examples or testing data is a waste of time IMHO. It happens regularly on these and other forums though. :talk028::salute:
 
J

jerryk

Guest
I've heard this since the G1 model released last November. Several months ago I decided to do some DR and meter calibration testing, comparing my E-1, L1, K10D, G1 ... and a week later had a D200 to put through the same test as well.

What I found was that the G1's ISO sensitivity setting is 0.3-0.5EV pessimistic compared to the E-1/L1/K10D and the D200's sensitivity about 0.5EV optimistic. What this means is that if you set ISO 100 on all five cameras, the actual sensitivity measured by a calibrated reference incident meter and gray card target indicates that the G1 sensor is actually at ISO 140-150, the Nikon is actually at about ISO 50-64, and the other three are with 5% of 100.

The next thing I noticed was that the metering characteristics of the cameras varied quite a bit too. The E-1 was most protective of highlight saturation (it would suggest about -0.7EV compared to a nominal reference reading), the L1 and K10D were closer to accurate at -0.3, the D200 showed nominal at 0, and the G1 tended to be right on the line or a little over at +0.3 EV.

These mapped pretty closely to my picture taking experience where I find I normally have the E-1 set to +0.7EV compensation, the K10D and L1 set to +0.3EV, the D200 at 0EV, and the G1 at -0.3 to achieve the same average mid-gray point with the RAW processor at the defaults.

Once I dialed in these corrections and ran the DR test, all five cameras showed remarkably consistent DR performance. Absolute numbers are not really very interesting, but relative performance is useful since they were all tested in the same circumstances using the same methodology .. DR decreases as ISO goes up, so the numbers are presented in stops as {lowest ISO} to {highest ISO}:

K10D: 10.5 to 9
L1, E-1: 10 to 8.5
G1: 10 to 7.5
D200: 9.5 to 8

So ... The G1's DR is fairly comparable to similar format DSLRs, with a little hit at the highest ISO settings. However, the combination of the most pessimistic ISO sensitivity setting and metering calibration that pushes the saturation exposure limit will artificially restrict DR if you don't adjust take the hot sensitivity and metering calibration into account.

The GF1 is the same sensor, same metering, and almost the same data system ... I'm sure the results for it would be very similar to the G1.

Could you explain how one does this test and derives these numbers? What is a calibrated incident meter? Will a Sekonics meter work?

I am just comparing the same object shot at the same in LR which is probably not the best way to objectively determine the relative DRs of my GF1 and 5DMK2.
 

Godfrey

Well-known member
Great post! You do the tests that I can't seem to find the time to do. Funny, I have had a GF1 and G1 for about a month and have noticed without testing what your tests show. I routinely dial down to -.3 to get the exposure that I want.
Thanks Chuck.

I don't often really enjoy testing, but I do it in the course of trying to understand my camera's behavior when I come up against problems or when friend/client/student of mine has a problem that isn't otherwise resolvable. This testing came about because one of my friends who owned the D200 was complaining about noise at ISO 400 and unusability at ISO 800 ... that didn't seem right to me, the D200 generation sensors should be perfectly useable at ISO 800 if the sensitivity and metering calibrations are in the right ballpark. So I set up to do the DR testing for that, and decided since I had the setup all ready to go it would only take me another half hour to test the other four cameras and get a baseline for comparison.

Based on the results of this testing, the problem was that his D200's ISO sensitivity calibration is optimistic and the meter is also protecting highlights too much, leading to almost an average full stop of underexposure which at ISO 400 and above induces a lot of noise. I used the D200's custom function to bias the metering calibration across the board, shifting it upwards by +0.7EV, and after that the camera produced much cleaner exposures: ISO 400 and 800 are now very usable and he's even told me that he can use ISO 1600 for some subject matter.

A little bit of good information can really make a difference. :)
 

Godfrey

Well-known member
Could you explain how one does this test and derives these numbers? What is a calibrated incident meter? Will a Sekonics meter work?

I am just comparing the same object shot at the same in LR which is probably not the best way to objectively determine the relative DRs of my GF1 and 5DMK2.
The DR testing setup is fairly tedious to write up ... it's a relatively simple extinction test in concept. You can find a very similar testing methodology published in Photo Techniques magazine, January-February 2007, Vol 28, #1 ... I do it just a little bit differently given the equipment at my disposal but the concept is the same. DR tests like this are good for relative comparison value between various bodies tested the same way, not absolute numbers.

My 'calibrated incident meter' is a Sekonic L328 FlashMate that I have calibrated with a known reference light source and against densitometer readings from film. It's very accurate and stable.
 
Top