Thu 06-May-10 12:31 AM | edited Thu 06-May-10 12:47 AM by eddie_kawasaki
I just recieved my new D90 and I just did a quick test shot compared with my D300. What am I doing wrong to make the pics look slightly different between the two cameras, the D300 looks darker to me. Both were shot with a Nikon 60mm 2.8 macro lens, at the same aperature, ISO, and shutter speed, with no flash. I took multiple shots with each camera to make sure its not the flourescent lights flickering through the cycles. Keep in mind that I'm at work with no tripod, just handheld sitting and leaning back into a chair to stabilize myself. The pic on top is from the D300.
According to the EXIF data, the exposures were made at 1/160 sec. Fluorescent lights cycle at 1/120 sec. (twice the power frequency). That means it is very likely that the two pictures were taken during a different portion of the lamp cycle. Try repeating the experiment with 1/60 or 1/30 sec. That way you will have an integer number of cycles of the lamp during the exposure.
Also as mentioned by esantos, white balance could be a problem. You used "auto". When you repeat the experiment at slower shutter settings, use the same fixed WB in each camera.
Gary in SE Michigan, USA. Nikonians membership - My most important photographic investment, after the camera. D4, D810, D300 (720nm IR conversion), D90, F6, FM3a (black), FM2n (chrome) YashicaMat 124, Graflex Speed Graphic 4x5 My Nikonians Gallery & Our Chapter Gallery
Could you perhaps explain why you did a comparison between the two cameras? If both cameras gave you identical shots how would you have felt? Being two different cameras I would expect there to be differences. Two different photographers using the same camera at the same settings can get different looking shots because a lot of the time it depends where you meter in the scene.
Except that the EXIF data for each photo suggests they were shot at the same shutter speed, aperture, and ISO. So, even if the Matrix metering is not the same between cameras, they got the same exposure value.
It looks like the D300 exposes somewhere between 1/3 and 1/2 stop less compared do the D90.
The important thing is never to stop questioning. -Albert Einstein
It is error only, and not truth, that shrinks from inquiry. -Thomas Paine
>Except that the EXIF data for each photo suggests they were >shot at the same shutter speed, aperture, and ISO. So, even if >the Matrix metering is not the same between cameras, they got >the same exposure value. > >It looks like the D300 exposes somewhere between 1/3 and 1/2 >stop less compared do the D90. > >Scott
Assuming that you are correct I don't see a problem. As I stated in another post they are different cameras and are bound to be different in exposing an image.
When doing this type of test, for whatever reason, it's necessary to eliminate as many variables as possible. The "normal" setting for ADL may not mean the same thing in different cameras, so ADL should be turned off. As already mentioned, a preset WB should be used instead of "auto". And if you're trying to assess the exposure the camera gives rather than its metering, turn off ISO Auto, switch to "M" mode, and set the aperture, shutter speed and ISO yourself.
I would also not shoot jpg. That is another big variable, as the raw conversion in-camera is likely not the same at all between the two.
Shoot raw, then use the same converter with the exact same settings.
To the comments "why compare" I guess I agree with the original poster -- I would not expect the shots to look identical at all, but if you do a completely manual exposure, I would expect the basic histogram to end up the same (not identical, but similar shape and placement). They are quite different, I tried to estimate in photoshop and got about a half stop. And they are still quite different if you correct the white balance first.
My guess is jpg conversion from raw is different.
Or maybe we all just presume that ISO really is well matched between cameras. I have never compared -- maybe they are always this far off. I'd be curious to see, say, two d300's.
The D90 is a newer camera than the D300. I don't think Nikon issued an update to Capture NX 1 to support the D90 and later cameras, so to process D90 NEF files directly you'd need to purchase a copy of Capture NX 2.
>The D90 is a newer camera than the D300. I don't think Nikon >issued an update to Capture NX 1 to support the D90 and later >cameras, so to process D90 NEF files directly you'd need to >purchase a copy of Capture NX 2.
Well, ViewNX is free, and I think is updated regularly, and will do a conversion if all you want is just a consistent conversion.
Sorry, I can't receive 12mb from e-mail, if you posted somewhere I could convert.
But ViewNX is handy anyway -- you might want to grab a copy. The price is right (it's a Nikon product also).
View NX is probably a better tool for comparison of RAW files. It will helpaccurately remove differences in post processing.
The other approach to comparison is to use the same picture control setting for raw images from each camera.
One thing to keep in mind is that even if you used the same picture control setting in NX2, you might find differences. Each camera is "tuned" to optimize images - and they are not necessarily tuned the same way. For example, the sensors are different so does matrix metering perform the same way? I expect they wil be close with the same camera settings, same picture control setting, same post processing, etc. - but they may not be exactly the same.
Is your goal to set up the two cameras to produce the same images? For certain situations creating the same type image from two cameraes would be useful. Or are you trying to compare and understand possible differences? Or is there some other goal?
It is possible in View NX to set an identical white balance to two images that are shot in RAW, but that won't resolve differences in metering or exposure.
Eric, I did not adjust anything with NX I only used NX last night to change the file size so I could post here. Now that you mention it, it would be nice to have both cameras set up to produce the same images since at times when shooting Taekwondo tournaments, when my son is either waiting or has already finished competing, he usually takes one of the cameras to the other side of the ring so that way we may be able to get a better or unobstructed view. Since I am selling the tournament photos having 2 cameras work the same fight can help me to produce and sell more. But this was not my goal, I was just playing around with the 2 cameras last night and noticed a difference, so I decided to ask here. No secrets or motives, just asking.
Ok, D lighting is now off, white balance was preset using an expodisc using sunlight this time as the source of light for the room,outside is clear skies no clouds. Does the white balance really matter with RAW? Well, I did the white balance preset already. Here are the newest photos, converted with View NX from RAW to Jpeg. This time with different lighting from last night, the photos are still darker with the D300. The D300 photo is the top one.
White balance should not affect shooting raw, though the white balance is read by most converters and does affect the conversion if you leave it set to "as shot" or anything other than a fixed amount.
But I agree, this does not look like white balance. TO me it looks like about half a stop (eyeball estimate) difference in exposure.
I really have no idea. It's not a test I have ever tried. When I get some time I have a D80 and D300 and can experiment. Off hand I would say one of your sensors is about half a stop more sensitive than the other even when set the same.
OK, silly question -- you don't have any EV+/- exposure adjustment set do you. That would override the actual exposure set.
BUt I come back to the original question -- I've never seen anything that says how accurate (vs. precise) the ISO setting is on a DSLR? I know it is quite precise, meaning I can vary it by fractions of a stop and make a difference. But how accurate is it -- how closely is it calibrated to some external metric.
Think of it like a scale that can weigh to fractions of a pound, but is skewed by a pound -- you can tell if the weight changes, but you might not be getting the right answer. How accurate is a D300 ISO scale supposed to be, relative to the "right" answer. Or a D90? I know that in my shooting, I would never notice - everything I do is relative to either an automatic metering, or where I adjust a bit up or down. Since the metering may be equally off (i.e. meaning compensating), I would never know if my ISO 200 is another person's ISO 300. Unless I did a test as you did. Very interesting.
>EV for both are at zero, last nights pics were with matrix >metering, todays pics are with centerweighted metering.
But I had the impression you were setting the exposure manually, so metering does not matter?
If you are not setting it manually, then this could explain it -- different metering on difference cameras, even between different firmware versions, certainly exists. I assume you are setting the exposure at a fixed ISO, f-stop, shutter speed.
yes, both cameras set to "M", same f stop, same shutter speed, smne ISO, and same metering mode. Setting exposure manually means to set the shutter speed, f stop, and ISO am I correct? Are you saying that there is an AUTO mode for Matrix, centerweighted and spot metering? If there is I am not using it.
If my memory serves me right there was a discussion in the Nikon Cafe last year about the differences between the d300 and the d300s regarding the same issue. Also between the d70 and the d80 sometime back. What Nikon does is listen to the buyers about different camera upgrades and if there are complaints about underexposure/overexposure then they tweak the sensor when a new camera is produced. Sometimes the complaints are user faults - I am not stating that you are guilty - and the tweaks aren't necessary. What you could do is to set the EV on the d300 up by 1/3 or 2/3 and repeat your experiment and look at the results. I wouldn't get too bogged down in this because you can always compensate if you want the two cameras to produce the same result. Post processing will do it very quickly even with a batch process.
According to the EXIF, I believe the first image uses Matrix metering (Opanda says it is Pattern) and the second is Center Weighted. That could account for the difference and is the most likely cause. Now since you are in Manual Mode, it should not really matter, but there could be other implications on the way the image is rendered.
I think this is a good discussion as there are plenty of times photographers want to set up two cameras the same way. It's important for many of us to understand the differences and how to manage them. One of the typical post processing routines is to apply the same white balance settings and Picture Control settings to all images.
Ok, now they are equal, I didnt realize that I forgot to set the metering the same for todays pics, but I just did 2 more and set both to center weighted. But the result is the same no matter, the D300 is still darker in appearance. D300 is the top photo
i think it comes down to picture control differences between cams even when converting nef files in nx the pic controls will still be different as nx still knows what cam it came from and applies it accordingly you need to convert the nef files in a non Nikon software that will take the pic controls out of the equation
You can view images in View NX without Picture control utility, at least in my copy. A non Nikon software program - such as ACR - will have to have the controls zeroed out. If picture control isn't switched on in camera then your post is pointless.
Brian if we accept that you can't switch off picture control and the picture controls between cameras are different then the posters dilemma can't be resolved? I have a linear setting on my camera - instead of neutral - that was uploaded to the camera which can be used in many types of cameras. This in effect switches off the picture control because it is an "universal" setting. Linear is in effect "no processing"
Fri 07-May-10 10:33 AM | edited Sat 08-May-10 09:43 AM by ericbowles
I think that is one of the key points here - that Neutral would not necessarily be exactly the same as the camera sensors are different and the way the camera's processor applies processing to the image would be different. I'm not talking about post processing - simply the way the camera interprets the light received by the sensor.
Probably an interesting exercise is to do what the originator of this thread suggests - try to set everything the same on multiple cameras - even the same camera model - and attempt to capture the same image and render it exactly the same way.
With past cameras Nikon tweaked the settigns for rendering RAW images. For example, the VIVID setting varies by model.
My expectation is that the same Neutral setting applied to images taken with a D90 and a D300 may produce slightly different images even if all the settings are the same.
Fri 07-May-10 10:26 AM | edited Fri 07-May-10 10:28 AM by Baaker
>You can view images in View NX without Picture control >utility
Does View NX automatically open raw files without picture controls? If not where do I find the setting to do so? I have the latest version that can be downloaded from Nikon unless the newest version does not have that option.
I have View NX 1.52. See help for version information. When you open it go to File ( top left hand corner) and you will see Launch Picture Control Utility. I don't see anywhere to launch the utility automatically when the program is opened.
Due to mfg.'s acceptable tolerances which we know can vary during manufacturing of not only the camera but its individual components isn't it possible that even from camera to camera of the same model for pictures to be slightly different? I mean if a component on one camera is within tolerances but more to the plus side and another camera of the same model is within tolerances but more to the minus side wouldn't the pictures look different? Or does the camera software take these into consideration. I was just curious as I have no idea or I could be just asking nonsense.
I don't think this is a matter of tolerances as you would not get much variation. I do know that Nikon tweaks the way a given setting works across models. They can also further adjust with firmware updates.
I find it very interesting that Eddie is able to demonstrate such a high degree of difference with what appears to be the same settings. The use of a single picture control setting to get the same result seems to point that the two cameras are different in the way they render the same image with the same settings.
I don't see that as a problem - just different, but if you are shooting an event with two cameras it could be problematic.
Looking at other camera models, the D2X had a specific way of rendering color that many find appealing. A Picture Control setting was created to render somthing close to that effect, but that does not mean they are exactly the same settings.
Unless someone can spot an inconsistency between Eddie's tow examples, we may have found a demonstration of a difference in the way two cameras render an image. It would be interesting if someone could repeat this test with the same two camera models or with other pairs of camera models.
Sat 08-May-10 04:19 PM | edited Sat 08-May-10 04:22 PM by WorldShooter
>I don't think this is a matter of tolerances as you would not get much variation.
I'm not so certain about this.
After all, in the D300 Custom Settings menu there is an option (#b6, ref page 277 of the English User's Manual) to "Fine Tune Optimal Exposure: Use this option to fine-tune the exposure value selected by the camera. Exposure can be fine tuned separately for each metering method by from +1 to –1EV in steps of 1/6EV."
Why would Nikon provide this option?
Apparently Nikon feels the optimal exposure can be off by a whopping +/- 1 EV as indicated by the camera's metering system! You can even optimize (in 1/6th steps) for each metering method (matrix, center-weighted, spot) separately.
The question remains off-course if an "optimal exposure" can be objectively determined. Or is it still a subjective value?
BTW1, I don't know how this works on the D90, as I don't have access to one or to its User's Manual at the moment.
BTW2, FWIW, to my layman's eyes the posted D300 pictures look fine, I would say the D90 pictures look somewhat over-exposed.
>So the pictures look the same no matter which pitcture >control setting I choose to open the raw file. > >Baaker, I have the same version, and there is no way to not >use a pic cntrl to open, you say that you can in YOUR version, >so detail how you did it.
Do a search for linear.ntc on the internet. Download it and it can be imported into the picture control utility in View NX. If you do a search you can also find out how to upload it to your camera. It is done via your card and it will appear in the menu in your camera. You can set it to be your default picture control. If you load it into two cameras then the images should have identical settings in respect to contrast etc etc.
I was curious, so I tried a D80 (note 80, not 90) and a D300.
I set up a tripod on a picture frame lit by artificial light (i.e. not changing), and set both cameras as follows:
- Raw - F4 - 1/40th - Manual - Auto-iso off - ISO 200
I then converted in Lightroom (3b2) using Adobe Standard (not the camera specific profiles), and did a reset (no presets, everything basically to zero). By using ACR I hoped to get away from any Nikon camera specific stuff.
I set white balance to a specific temperature (tungston preset) so it would be the same interpretation on each.
BIG difference not just in overall tone, but in exposure.
It's about 2/3rds of a stop. What might I have missed?
Note I am not surprised of tone difference, white balance, etc. But I am surprised at that degree of exposure difference.
For what it is worth my wife and I have three D300's - his, her and a spare body. With all settings identical and viewing through either NX2 or CS4 ADR, one body is 1/2 stop brighter than the other two identical bodies. Not a problem for us. We just adjust accordingly to get the same images from each camera. Jerry R
I'm at work and don't have either camera with me to experiment so I can't answer this question. If I bump the EV up a stop will it affect my shutter speed? I think it does but I don't remember. As most of my shooting is in dimly lit high school gyms, I need as much shutter speed as I can get to freeze the actions of kicks at Taekwondo tournaments.
>If I bump the EV up a stop will it affect my shutter speed? >I think it does but I don't remember.
Changing exposure comp adjusts the meter's reading.
What the camera does with this modified meter reading depends on what mode you're shooting (PASM) and if you're using auto ISO. In aperture priority with auto-ISO off, chaging the meter's reading will directly modify shutter. If auto-ISO is on, it could modify shutter and/or ISO. In manual mode, it effects only the meter readout in the viewfinder. Not sure about manual + auto-ISO.