Two cameras, F16, ISO 400, 1/125th - different results

FlyTVR

FlyTVR

Messages
4,234
Edit My Images
No
I know I am missing something fundamental.

Lets say (under studio conditions) you set up your lighting / camera for the correct expose e.g. F16 ISO 400 1/125th.

Should you then be able to pick up a different camera / lens with the same settings (F16 ISO 400 1/125th) and get a similar exposure?

Your thoughts please.

Cheers,

Dav
 
Were the two cameras the same model or different ?

Also when you say the settings were the same - do you mean all settings or just those 3 you've mentioned ?
 
dinners said:
Were the two cameras the same model or different ?

Also when you say the settings were the same - do you mean all settings or just those 3 you've mentioned ?
Click to expand...

Camera A - Canon 5DII, Sigma 24-70 F2.8
Camera B - Canon 7D, Canon 70-200 F2.8 MKII

What other settings are you thinking about Phil?

Cheers.

Dav
 
It was the metering settings really that sprung to mind. Also exposure compensation.

Not sure if the focal length might have contributed ?

70mm is as long as you could have been on the full frame 5D2
70mm x 1.6 = 112mm is as short as you could have been on the 7D

Sensor size may have an effect and the different lens element arrangements can create variables.
 
Last edited:
How much out are you? Just because the lens is told f16 doesn't mean it will be exactly f16. Same for ISO and shutter speed (but I'd assume those to be closer to correct values).
 
arad85 said:
How much out are you? Just because the lens is told f16 doesn't mean it will be exactly f16. Same for ISO and shutter speed (but I'd assume those to be closer to correct values).
Click to expand...

Totally agree, BUT -

Light meter - 308S F16
F16 on the 5D
F22 on the 7D

This was the first time I've ever used the 7D on a job - so I was..... left :thinking:

I must be missing something obvious!!!

Cheers.

Dav
 
Is the result the same if the lenses are swapped?
 
Trouble is, this makes the use of a light meter pointless.

In my little head, if the light meter says F16 at ISO 400, that's what my camera should be set at! In the case of the 5DII+Sigma that is true, but the 7D+Canon......

Cheers.

Dav
 
FlyTVR said:
Totally agree, BUT -

Light meter - 308S F16
F16 on the 5D
F22 on the 7D

This was the first time I've ever used the 7D on a job - so I was..... left :thinking:

I must be missing something obvious!!!

Cheers.

Dav
Click to expand...


I read something recently about tests that DXO had carried out regarding ISO settings and their conclusion was that even between cameras from the same manufacturer there were noticeable differences in what one camera said was a particular value and what another model considered it to be.
 
Flash In The Pan said:
I read something recently about tests that DXO had carried out regarding ISO settings and their conclusion was that even between cameras from the same manufacturer there were noticeable differences in what one camera said was a particular value and what another model considered it to be.
Click to expand...

Great - this gets better! :LOL:

Dav
 
you input one value if i am not mistaken, probably a shutter speed in studio enviroment, the meter will take a reading know the shutter speed required and work out the aperture for a correct exposure.
 
33L said:
you input one value if i am not mistaken, probably a shutter speed in studio enviroment, the meter will take a reading know the shutter speed required and work out the aperture for a correct exposure.
Click to expand...

You specify shutter speed and ISO. It then tells you the F-stop

Looking at it, that's only one (full) stop out. Correct?

Having said that, sort of blows away the purpose of a light meter!
 
Last edited:
I seen to recall the same thing concerning ISO as FITP. But to answer your question, in theory yes you should be able to pick up a different lens/camera combination and transfer the settings from the light meter and get a similar exposure. Did you try setting the light meter readings into each camera and compare the actual exposures?
 
Last edited:
The element that is 'wrong' here is the 7D isn't it? If the exposure is fine with the meter and the 5D and yes, all things being equal you should just transfer the meter reading to the camera, end off. Have you got any exposure compensation set on the 7D?

Paul
 
paul williams said:
The element that is 'wrong' here is the 7D isn't it? If the exposure is fine with the meter and the 5D and yes, all things being equal you should just transfer the meter reading to the camera, end off. Have you got any exposure compensation set on the 7D?

Paul
Click to expand...

Thanks Paul.

Shot fully manual - no ev

Finger might be pointing at the 7D, but I can't rule out the 70-200 as I didn't have time to start running experiments. :(
 
Google "7D exposing differently"
 
Okay, short answer, you calibrate your meter to your camera. Some of the higher end sekonics let you save a few different cameras.

Longer answer, well, ISO shouldn't vary but it does, quite a bit
Also welcome to the concept of T-stops which for some reason is only really used in the film world but applies to photographers as well.

Basicly your aperture setting on each lenses assumes no loss of light moving through the glass... this is near enough impossible! t stops is how much light is lost by light moving through the lens and of course varies between them.

I just use a basic sekonic meter and know my camera ties up with it fairly well iso wise and my lenses seem to be close enough to each other, but i did test with a grey card and will test with future cameras and lenses.
 
nano said:
Okay, short answer, you calibrate your meter to your camera. Some of the higher end sekonics let you save a few different cameras.

Longer answer, well, ISO shouldn't vary but it does, quite a bit
Also welcome to the concept of T-stops which for some reason is only really used in the film world but applies to photographers as well.

Basicly your aperture setting on each lenses assumes no loss of light moving through the glass... this is near enough impossible! t stops is how much light is lost by light moving through the lens and of course varies between them.

I just use a basic sekonic meter and know my camera ties up with it fairly well iso wise and my lenses seem to be close enough to each other, but i did test with a grey card and will test with future cameras and lenses.
Click to expand...

Thanks for that :) Nice to know I wasn't going mad!
 
FlyTVR said:
Totally agree, BUT -

Light meter - 308S F16
F16 on the 5D
F22 on the 7D

This was the first time I've ever used the 7D on a job - so I was..... left :thinking:

I must be missing something obvious!!!

Cheers.

Dav
Click to expand...

I don't understand what you are saying here. If the flash meter says f/16 why have you set the 7D to f/22? The 7D cannot meter flash and neither can the 5D, so the camera metering is irrelevant. Are you saying that you need to stop the 7D down by a full stop in order to match the exposures visually between the two cameras?

Do you have ALO, LPIC or HTP enabled on the 7D? Same picture styles on both?
 
tdodd said:
I don't understand what you are saying here. If the flash meter says f/16 why have you set the 7D to f/22? The 7D cannot meter flash and neither can the 5D, so the camera metering is irrelevant. Are you saying that you need to stop the 7D down by a full stop in order to match the exposures visually between the two cameras?

Do you have ALO, LPIC or HTP enabled on the 7D? Same picture styles on both?
Click to expand...

Yes
 
Tim's question about camera settings needs to be answered, could be the difference. Otherwise the T-stop is another factor, but I suspect the other one is f/16. It's a very small aperture, and sometimes prone to mechanical inaccuracy, ie it's not actually stopping down to exactly the right f/number it should be.

I think you'd be unlucky to be that far out with quality gear like that, so assuming nothing is actually faulty then it's probably a compound of all these factors, maybe double compounded by the other camera being a bit the other way - if you see what I mean. When you look at all the possible variables, and if they all run against you, it's easy to account for a one stop variance.

But you're right about hand-meters being at a disadvantage sometimes. They make assumptions about what happens to the light after it enters the lens, sometimes incorrect ones, and they know nothing about things like T-stops and vignetting which a TTL meter will automatically compensate for. Which is why I always use the LCD/histogram/blinkies when it comes to precise exposure setting - ultimately it doesn't matter about anything else, what appears on the sensor is what matters, and that's what they show.
 
using T stop instead is of F stop is part of the answer

a light meter you can calibrate to each camera (e.g. 758D) is another

a spot meter instead of an incident meter is also worth considering, as that will take into account the reflectivity of the subject.

there's lots of ways of approaching the matter...

...but nothing electronic beats taking a picture, looking at the results, and adjusting either the lighting or the camera based on this information!
 
tdodd said:
I don't understand what you are saying here. If the flash meter says f/16 why have you set the 7D to f/22? The 7D cannot meter flash and neither can the 5D, so the camera metering is irrelevant. Are you saying that you need to stop the 7D down by a full stop in order to match the exposures visually between the two cameras?

Do you have ALO, LPIC or HTP enabled on the 7D? Same picture styles on both?
Click to expand...

I've only just purchased the 7D, so I'll have to fess up to the fact that I don't know. It would help me (and possibly others?) if you could expand on ALO, LPIC and HTP.

Cheers.

Dav
 
FlyTVR said:
I've only just purchased the 7D, so I'll have to fess up to the fact that I don't know. It would help me (and possibly others?) if you could expand on ALO, LPIC and HTP.

Cheers.

Dav
Click to expand...

He means the in-camera processing pre-sets, in Custom Function II/Image and Picture Styles.

They all affect the JPEG image exposure, so must be the same. In particular, the Contrast setting in Picture Styles changes things quite a bit (see how it shifts the histogram) and Highlight Tone Priority also impacts the Raw (by effectively reducing ISO in bright areas).

Edit: when you say the exposures were different, what are you comparing? The LCD, histograms, monitor image, prints?
 
Last edited:
I know nothing about Canon but i know that if one Nikon had ' active D lighting ' turned on & the other didn't the results would be very different. Does Canon have anything like ' D lighting '?

Or maybe the picture profiles used were very different on each camera.
 
ALO - Automatic Lighting Optimisation is, I suspect, Canon's equivalent of D-Lighting. I think it is enabled by default. I routinely turn off as much in camera processing as soon as I get a new camera. If anyone or anything is to fiddle with my images it will be me, not the camera. How anyone can judge their exposure accuracy when the camera is actively adjusting things is beyond me.

HTP - Highlight Tone Priority - underexposes by a stop by "secretly" lowering the ISO (not a well kept secret) and then applying a special tone curve to boost the shadow areas while preserving extra highlight headroom. You get more shadow noise as a result. If you like to shoot ETTR then using HTP is self defeating.

LPIC - Lens Peripheral Illumination Correction - compensates for corner falloff on a lens by lens basis. On a 7D it should have less effect than on full frame since the corners aren't used. You may need to load different lens profiles into the camera from EOS Utility if the standard set within the camera doesn't include your lens. I don't know if the feature will cater for non Canon lenses. Since I only have Canon glass and don't use the feature anyway I neither know or care.

Picture Styles - not a new concept. They've been included in cameras since the 30D and 5D. The effect of various options should be well known. Shoot raw and play in DPP if you want to explore the differing results from each picture style.

I think the comments about T-Stops and sticky aperture blades should not be ignored.
 
tdodd said:
ALO - Automatic Lighting Optimisation is, I suspect, Canon's equivalent of D-Lighting. I think it is enabled by default. I routinely turn off as much in camera processing as soon as I get a new camera. If anyone or anything is to fiddle with my images it will be me, not the camera. How anyone can judge their exposure accuracy when the camera is actively adjusting things is beyond me.

HTP - Highlight Tone Priority - underexposes by a stop by "secretly" lowering the ISO (not a well kept secret) and then applying a special tone curve to boost the shadow areas while preserving extra highlight headroom. You get more shadow noise as a result. If you like to shoot ETTR then using HTP is self defeating.

LPIC - Lens Peripheral Illumination Correction - compensates for corner falloff on a lens by lens basis. On a 7D it should have less effect than on full frame since the corners aren't used. You may need to load different lens profiles into the camera from EOS Utility if the standard set within the camera doesn't include your lens. I don't know if the feature will cater for non Canon lenses. Since I only have Canon glass and don't use the feature anyway I neither know or care.

Picture Styles - not a new concept. They've been included in cameras since the 30D and 5D. The effect of various options should be well known. Shoot raw and play in DPP if you want to explore the differing results from each picture style.

I think the comments about T-Stops and sticky aperture blades should not be ignored.
Click to expand...

Off topic ;)

Is that your experience with HTP Tim?

My own findings, admittedly hardly exhaustive, is that with Highlight Tone Priority there is definitely about a stop more highlight detail, but minimal impact (that I can see) on shadow noise. A net upside, and this advantage is recorded in the Raw. Shouldn't that make it a better starting point for any form of Raw post processing, if maximum dynamic range is the objective, eg ETTR?

Is there any difference between applying special shadow tone curves, and shooting at a higher ISO in the first place, in terms of noise, given that ISO is basically just sensor gain anyway? I'm thinking that since exactly what Canon is actually doing with HTP is only known to them, and assuming they know more about processing their Raws than say Adobe with their multi-platform profiles, isn't is reasonable to assume that what Canon gets out of the Raws in this way is likely to be better than we can get by tweaking things ourselves in Lightroom?

I'd be interested in your comments/findings. Thanks :)
 
My experience of HTP is that I don't use it. I used it once, in ignorance, back in Jan 2008 and never again since. No harm to the images since they were snow scenes and very, very bright. But I am now better informed and would not dream of using it, AS A RAW SHOOTER. If you shoot to JPEG then that's a different matter, but one which does not apply to me.

Those who have analysed RAW files produced with and without HTP enabled have determined that the RAW data from an image shot at 200 ISO with HTP is the same as the raw data from the same scene shot at 100 ISO without HTP. HTP "buys" you more highlight headroom by applying less gain to the sensor output. You end up metering for one ISO value and shooting at another. You are moving your exposure one stop to the left.

While the raw data will appear the same from either approach, with HTP enabled you will get an entirely different image and histogram on the back of the camera. You will be seeing the results of gain applied by the tone curve instead of gain applied to data from the sensor. If your goal AS A RAW SHOOTER is to expose to the right then HTP will undermine your efforts by showing you lies on the back of the camera. If you shoot to JPEG then no problem. What you see is what you get. With raw, what you see is not what you get, not at all.

Richard, you know full well the impact of picture styles and contrast settings on the "accuracy" of the histogram for raw shooters. Well add HTP to the list of conspirators. You can probably put it at the top of the list.

If HTP is so good, why is it even an option? Why not make it part of the standard processing? Because it robs Peter to pay Paul. You can't magic up extra dynamic range from the sensor. You can only fiddle with your processing of that data. If you want extra highlight headroom you have to sacrifice shadows. Canon knows that. The warnings about increased shadow noise are there in the manual.

I won't say that there are no advantages to HTP, because for some people I know there are, but for me, as a raw shooter who meters for highlights and does not use DPP there are none. Absolutely none. If you're shooting to JPEG on the fly and aren't in a position to nail your exposures every time then it may well be useful. In the studio, with controlled lighting and metering, no way, even with JPEG, unless you happen to especially like the tone curve it employs.
 
tdodd said:
My experience of HTP is that I don't use it. I used it once, in ignorance, back in Jan 2008 and never again since. No harm to the images since they were snow scenes and very, very bright. But I am now better informed and would not dream of using it, AS A RAW SHOOTER. If you shoot to JPEG then that's a different matter, but one which does not apply to me.

Those who have analysed RAW files produced with and without HTP enabled have determined that the RAW data from an image shot at 200 ISO with HTP is the same as the raw data from the same scene shot at 100 ISO without HTP. HTP "buys" you more highlight headroom by applying less gain to the sensor output. You end up metering for one ISO value and shooting at another. You are moving your exposure one stop to the left.

While the raw data will appear the same from either approach, with HTP enabled you will get an entirely different image and histogram on the back of the camera. You will be seeing the results of gain applied by the tone curve instead of gain applied to data from the sensor. If your goal AS A RAW SHOOTER is to expose to the right then HTP will undermine your efforts by showing you lies on the back of the camera. If you shoot to JPEG then no problem. What you see is what you get. With raw, what you see is not what you get, not at all.

Richard, you know full well the impact of picture styles and contrast settings on the "accuracy" of the histogram for raw shooters. Well add HTP to the list of conspirators. You can probably put it at the top of the list.

If HTP is so good, why is it even an option? Why not make it part of the standard processing? Because it robs Peter to pay Paul. You can't magic up extra dynamic range from the sensor. You can only fiddle with your processing of that data. If you want extra highlight headroom you have to sacrifice shadows. Canon knows that. The warnings about increased shadow noise are there in the manual.

I won't say that there are no advantages to HTP, because for some people I know there are, but for me, as a raw shooter who meters for highlights and does not use DPP there are none. Absolutely none. If you're shooting to JPEG on the fly and aren't in a position to nail your exposures every time then it may well be useful. In the studio, with controlled lighting and metering, no way, even with JPEG, unless you happen to especially like the tone curve it employs.
Click to expand...

Great reply (y) Thank you Tim.

I was googling around on this and came across some apparently authoritative comments (though they seemed a few years old, eg Bob Atkins) who was saying that with HTP the Raw was in fact no longer Raw in the commonly understood form. Perhaps what I'm seeing is camera pre-sets automatically reapplied as the Raw is opened. Or something. I need to do some more but since there is so little noise with the 5D2 anyway it's not easy to tell!

This is the bit I was wondering about "You will be seeing the results of gain applied by the tone curve instead of gain applied to data from the sensor." I assumed what actually happened was the scene was shot at the set ISO, and it was just the highlight end that was held back. And further, in practical terms what is the difference between gain applied to the tone curve and raising ISO?

Sorry this is OT. If you have any links that come to mind, that would be great. Cheers :)
 
Do a search for articles (forum posts) written by a chap called Gabor (GaborSch) and who also used the handle "Panopeeper". Sadly he is no longer with us, but he developed and maintained the Rawnalyze software, which can analyse unprocessed raw data and produce actual raw histograms. He has written on many sites including DPReview and POTN. There are many who would agree with his conclusions that HTP simply shoots with the ISO dropped by a stop, thus "underxposing" the data capture, and then bringing shadows and midtones back to where they belong while rolling in the highlights more gently.

p.s. I'm not sufficiently up on the technical whys and wherefores, but as you raise ISO you reduce read noise from the sensor. Quite how that works I have no idea, but the extrapolation from that is that you are better off raising the ISO when shooting than pushing an underexposed file in post. i.e. you are better off shooting at 200 ISO for a correct exposure to begin with than to shoot at 100 ISO and then apply a fancy tone curve to bring the shadows and midtones up to snuff. The same argument applies at least as far as 1600 ISO. Beyond that, depending on the camera in question, there may be no advantages to pushing the ISO higher, since in some cameras ISOs above 1600 are faked anyway and create their own problems. It is perhaps worth noting, in the 40D for example, that with HTP enabled you are unable to select either ISOs below 200 or above 1600. This is because the camera reserves those ranges for its own use (i.e cheating/faking).
 
Last edited:
My guess is you should be able to get the same result as long as you match the focal lengths.
 
I have got to send my 24-70 to Sigma as it constantly under exposes by 1.5 stops, sigma have confirmed they have seen this before. Just for your info.
 
tdodd said:
Do a search for articles (forum posts) written by a chap called Gabor (GaborSch) and who also used the handle "Panopeeper". Sadly he is no longer with us, but he developed and maintained the Rawnalyze software, which can analyse unprocessed raw data and produce actual raw histograms. He has written on many sites including DPReview and POTN. There are many who would agree with his conclusions that HTP simply shoots with the ISO dropped by a stop, thus "underxposing" the data capture, and then bringing shadows and midtones back to where they belong while rolling in the highlights more gently.

p.s. I'm not sufficiently up on the technical whys and wherefores, but as you raise ISO you reduce read noise from the sensor. Quite how that works I have no idea, but the extrapolation from that is that you are better off raising the ISO when shooting than pushing an underexposed file in post. i.e. you are better off shooting at 200 ISO for a correct exposure to begin with than to shoot at 100 ISO and then apply a fancy tone curve to bring the shadows and midtones up to snuff. The same argument applies at least as far as 1600 ISO. Beyond that, depending on the camera in question, there may be no advantages to pushing the ISO higher, since in some cameras ISOs above 1600 are faked anyway and create their own problems. It is perhaps worth noting, in the 40D for example, that with HTP enabled you are unable to select either ISOs below 200 or above 1600. This is because the camera reserves those ranges for its own use (i.e cheating/faking).
Click to expand...

Thanks for you time Tim. I'll have hunt around :)

Edit: found loads of stuff, cheers. Turning HTP off... though I do like it for JPEGs :D
 
Last edited:
You must log in or register to reply here.
Back
Top