Yet more on DA16-45 & exposure

[malfunction]

Hi All,

Just been away for a few days in Connemara in the west of Ireland, "testing" my DS with my DA 16-45 and FA 50 lenses, and noticed (among all the excellent scenery), the same issue others have been reporting with the DA lens.

I know there have been a good few threads on this issue, but I wasn't getting a clear message on what issue if any there actually is with the lens.

Anyway, here is one typical image in its original form, the same image with the sensitivity upped by one stop, and the histogram of the original image.







You can see the original image is very dark. The conditions were a little overcast, but nothing too nasty. You can see that the histogram is bunched way over to the left hand side. However, none of the channels are clipped in any way. Upping the sensitivity by 1 stop in Pentax photo lab gave a much more balanced image.

This histogram is a typical (if slightly exaggerated) example of the shots taken with the DA 16-45. By contrast, shots taken with the FA50 were always perfectly centred, with no clipping of channels either

My question is, is this really an underexposure problem? Adding 1 stop to all my exposures with this lens is easily doable, but I'm likely to forget this sometimes when I'm using other lenses so it's an irritation.

Alternatively, is this an issue of contrast rather than exposure? The way the histogram is all bunched up compared to the FA 50 (on the same subject) suggests to me that the DA isn't capturing the same range of brightness levels as the FA 50.

I know I can stretch this in Photo Laboratory or other software, but then you're interpolating information which isn't in the original image.

Does anyone have any idea of cause, effect and solution to this? Preferably an in-camera solution. I'm not trying to start another flaming session here, the lens produces otherwise excellent images. I'm just trying to find the best way to use the lens and camera to produce good shots..

Thanks for your help,
Mal..

PS

I'll be posting some pretty pictures here, just as soon as I get them suitably tarted-up. Err, I mean "post-processed"....
--
http://www.pbase.com/malfunction/galleries

 

[hugnut]

Of course, this problem has been reported by various users(mostly D users) for over a year now, I guess you would either live with having to correct or compensate pictures from the DA 16-45 while taking the picture or afterwards in editing software or sell it for another model/brand lens. I sold mine after 4 weeks and have tried various other lenses since. I currently have the FA 20-35 that exposes properly for me. It is kinda strange when someone reports this happening on their DA16-45 that everyone else who may not be having the problem "jumps" all over them as if they are making it up or doing something wrong, luckily it's just words and not stones

Hi All,

Just been away for a few days in Connemara in the west of Ireland,
"testing" my DS with my DA 16-45 and FA 50 lenses, and noticed
(among all the excellent scenery), the same issue others have been
reporting with the DA lens.

I know there have been a good few threads on this issue, but I
wasn't getting a clear message on what issue if any there actually
is with the lens.

Anyway, here is one typical image in its original form, the same
image with the sensitivity upped by one stop, and the histogram of
the original image.







You can see the original image is very dark. The conditions were a
little overcast, but nothing too nasty. You can see that the
histogram is bunched way over to the left hand side. However, none
of the channels are clipped in any way. Upping the sensitivity by 1
stop in Pentax photo lab gave a much more balanced image.

This histogram is a typical (if slightly exaggerated) example of
the shots taken with the DA 16-45. By contrast, shots taken with
the FA50 were always perfectly centred, with no clipping of
channels either

My question is, is this really an underexposure problem? Adding 1
stop to all my exposures with this lens is easily doable, but I'm
likely to forget this sometimes when I'm using other lenses so it's
an irritation.

Alternatively, is this an issue of contrast rather than exposure?
The way the histogram is all bunched up compared to the FA 50 (on
the same subject) suggests to me that the DA isn't capturing the
same range of brightness levels as the FA 50.

I know I can stretch this in Photo Laboratory or other software,
but then you're interpolating information which isn't in the
original image.

Does anyone have any idea of cause, effect and solution to this?
Preferably an in-camera solution. I'm not trying to start another
flaming session here, the lens produces otherwise excellent images.
I'm just trying to find the best way to use the lens and camera to
produce good shots..

Thanks for your help,
Mal..

PS
I'll be posting some pretty pictures here, just as soon as I get
them suitably tarted-up. Err, I mean "post-processed"....
--
http://www.pbase.com/malfunction/galleries

--
Thanks,
Hugnut

 

[jon9]

I am curious, have you set your camera to spot metering and taken comparitive photos with the spot aimed at exactly the same place and then compared exposures for the two lenses? Also, has either of your two photo's been cropped, if so, a full picture may well have different light content when full metering is used for both lenses.

I'll admit to haveing some exposure surprises with my Ds, but when it appears to err, it's always correctable, though I'd think I shouldn't have to.

just novice thoughts....
--
Jon

 

[mvb123]

I have no real explanation for the underexposure of the 16-45, indeed it often needs +0.3/+0.5 compared to older lenses. But I'm not really bothered by this because of the procedure I (often) follow. I look at the histogram of the camera to make sure that it has only a small flat line at the right side, largely following the 'expose to the right' advise. Of course not if there's nothing to be bright in the shot, and if there is white without really any detail then I don't mind if that starts highlight-blinking on the review image. Obviously blown highlights are to be avoided, but if you underexpose too much you're also losing data because the image-chip can distinguish more different levels in the bright-tones than in the darker-tones. So I often have to change the exposure, even with other lenses than the 16-45.

I don't know what metering mode you use, but my guess is that the image you showed needs a small positive correction and on top of that extra correction is needed because of the 16-45 lens.

Just my 2 eurocents..
--
Menno

 

[malfunction]

Both images are the same shot, uncropped but resized, one with a 1-stop brightening. I didn't post a 50mm image here.

I don't mind doing the post processing, but I'd rather the shot was as well exposed as possible at the outset.

Mal..
--
http://www.pbase.com/malfunction/galleries

 

[mikiw]

Hi,

I have the same underexposure problems with my 16-45, as I recently rported and been answered in-
http://forums.dpreview.com/forums/read.asp?forum=1036&message=13271995

It seems that you need another +0.5 stop correction, your corrected image is too dark, I think, as I have found that typically +1.5 is the right compensation.

Actually, with all other lenses I have, even primes, I have found that typically at least 0.5 stop is needed.

What I dont understand, is why this phenomena is lens dependent? the exposure meters does it readings TTL, so even if the lens attenuates the light, the sensor feels it. It seems that Pentax decided to something extra in the immage processing, which depends on the type of lens.

I also think that it is annoying, though cureable.

But, I am not sure that we have to blame Pentax. The Dynamic range (latitude) of the CCD is quite limited in comparison to negatives, especially when you shhot JPEG which compress it to 8 bits (255) from the sensor 12 bits (thats why it is more nicely compensated from RAW).

Actually, if you look at a photo lab operator's work, you'll notice that on almost all frames he does brightness and contrast corrections. CCD DR is more like slides - how many good slides you used to get from a roll with AE shots? 15? 20? only when you shot manually, with a lot of carefullness on the readings within the frame, u probably got more.

I think that Pentax took a more conservative approch. Over exposure is much worse than underexposure, blown out area canoot be corrected, and they are more annoying to the eyes than dark areas where you couldn't distiguish all the shades of black on a perfect lowtones exposure.

I think that Pentax thought that someone willing to spend 1500$ on a body +16-45 zoom, will not mind the extra work of postprocessing, which was actually done in the negatives days by the lady/guy in the photolab.

The real question is if Nikon and Canon do a better job in out of the camera exposures.

What I heard, is the shot from N and C looks better exposed and contrasty, but that they often overexpose on bright spots.

If this is the case (any of you guys ahve personnal experience with N or C?) - I preffer the Pentax conservative approch.
Miki

 

[macdaddy1]

Having just dumped a Canon Rebel DSLR with 3 very good zooms, I can say from 19 months with it that it ALWAYS overexposed the bright areas and tended to blow them out! I much prefer the small amount of post processing necessary with my *ist DS than have an image that either requires huge amounts of post processing or may even be unrecoverable.

My thoughts are that Pentax has made a superb camera and (most) lenses that require photographers to take control of the post process, much as if you had your own darkroom to do it with film.

Which would you rather have—the dumbed down automation that C & N force on you at this price point or the photographic freedoms Pentax allows for your creativity? I'll choose the freedom to "tweak" my shots, thank you!
Rob "MacDaddy" White

 

[Joele]

Having just dumped a Canon Rebel DSLR with 3 very good zooms, I can
say from 19 months with it that it ALWAYS overexposed the bright
areas and tended to blow them out! I much prefer the small amount
of post processing necessary with my *ist DS than have an image
that either requires huge amounts of post processing or may even be
unrecoverable.

Same experience with my friends 20D, I would much prefer an occasional slight underexposure, as compared to constant blown highlights. Your much more likely to be able to correct a slight underexposure in my exerience...

------------
Joel - *ist DS/P30n/ME
http://www.pbase.com/joele

 

[Brett St Pierre]

Actually, if you look at a photo lab operator's work, you'll notice
that on almost all frames he does brightness and contrast
corrections. CCD DR is more like slides - how many good slides you
used to get from a roll with AE shots? 15? 20? only when you shot
manually, with a lot of carefullness on the readings within the
frame, u probably got more.

Your right, people who have only shot print film all their life and not devoted a lot of time to their exposure don't realise how the photo labs have been covering for them all these years! Then they get a shot that couldn't be saved in the lab and they wonder what went wrong......

The real question is if Nikon and Canon do a better job in out of
the camera exposures.
What I heard, is the shot from N and C looks better exposed and
contrasty, but that they often overexpose on bright spots.
If this is the case (any of you guys ahve personnal experience with
N or C?) - I preffer the Pentax conservative approch.

That has been the consensus for quite a while on this forum from those that have used both, so could be the trusth.

--
Brett



The Journey is the Thing

 

[mikiw]

Thanks,

That what I thought. There can't be miracels. As I said, I prefer underexposure and postprocessing that results with excellent pics.
Miki

 

[Dejvid Zaninovic]

+1 to +1.5 compensation seems about right for my D and 16-45. I set +1 compensation in camera and set additional compensation when processing raw if needed, which is usually additional +0.5.

 

[Dejvid Zaninovic]

I just checked with IT8 target. For 16-45 I need +1.0 comp. when using no flash, when I use flash I need +1.5. Probably the camera is more afraid of blown highlights when using flash. With Tamron SP90 and Pentax F70-210 I need +0.5 without flash and +1.0 with flash. Now, if there was a way to configure this in firmware so I don't have to remember to set it.

 

[hugnut]

How about when you use a Pentax prime?

I just checked with IT8 target. For 16-45 I need +1.0 comp. when
using no flash, when I use flash I need +1.5. Probably the camera
is more afraid of blown highlights when using flash. With Tamron
SP90 and Pentax F70-210 I need +0.5 without flash and +1.0 with
flash. Now, if there was a way to configure this in firmware so I
don't have to remember to set it.

--
Thanks,
Hugnut

 

[Dejvid Zaninovic]

I don't have Pentax primes any more, too much hassle, there were all SMC-M. I will probably leave F70-210 and SP90 without compensation as +0.5 can easily be fixed but 16-45 really needs to be compensated, it is too big a difference.

I just checked with IT8 target. For 16-45 I need +1.0 comp. when
using no flash, when I use flash I need +1.5. Probably the camera
is more afraid of blown highlights when using flash. With Tamron
SP90 and Pentax F70-210 I need +0.5 without flash and +1.0 with
flash. Now, if there was a way to configure this in firmware so I
don't have to remember to set it.

--
Thanks,
Hugnut

 

[Lance B]

I did my own test this morning with the DA 16-45 and the SMC A 28mm f2.8 on the *ist D. The results are as below. All shot RAW and converted in C1 with absolutely no post process, just converted to tiff and then to jpeg for web view. It was shot through a dirty window, so there maybe a bit of blurriness but this is for exposure accuracy only and not for lens sharpness.

The results were all basically the same. I also tested the Tamron 28-75 later @ 28mm and it was close to the 16-45 but not enough in it to call.

As far as I could see the histogram for the 16-45 was perfect in the examples I tried, but the A 28mm was overexposing quite a bit. The histogram for the A 28mm was hard up against the right and blown, whereas the 16-45 was spot on.

The main culprit is the individual colour channels and particularly blue in this case. The 16-45 had a small blue spike perfectly as far right as could be but not overexposed, whereas the 28mm blew the blue(sorry about that) and a touch of green.

This is not a definitive test and I will try some more test pics at a later date.
DA16-45, ISO200, 1/200, f8



SMC A 28mm f2.8 ISO200, 1/200, f8



Comments welcome.
--
Lance B

 

[Mark Ransom]

I don't need a histogram to tell me that the A28 picture is over exposed. How can this happen when both lenses used the same f-stop and shutter speed?

What metering method did you use? Perhaps with a different metering method, the A28 would have been spot on and the 16-45 would have been under exposed.

 

[itev]

Photo with Pentax 16-45 :

Colors, light and contrast have not been modified



--
itev

*istD

 

[Joe Franek]

Dejvid Zaninovic wrote:
I don't have Pentax primes any more, too much hassle, there were
all SMC-M. I will probably leave F70-210 and SP90 without
compensation as +0.5 can easily be fixed but 16-45 really needs to
be compensated, it is too big a difference.

I wouldn't compensate my 16-45. Although most of my images are underexposed, sometimes they're right, and even overexposed.

this one had to be pulled back in CR to get the detail in the whites. (BTW If anything, I would expect this shot to be underexposed because of the whites.) -



this one is about right -



this one is not bad as is, maybe tiny bit under -



I think it would be a mistake to permanently compensate my 16-45. And to do it on a case-by-case basis is a lot of hassle. I rather enjoy the scenery like that beautifull bride and do the compensation in PS.

But of course, it is possible that some 16-45's underexpose more than others.

Joe

 

[Lance B]

I don't need a histogram to tell me that the A28 picture is over
exposed. How can this happen when both lenses used the same f-stop
and shutter speed?

I'm not sure why the difference, but the DA16-45 is spot on. By the way, all 3 lenses tested, DA16-45, A 28mm f2.8 and the Tamron 28-75 F2.8, used the same shutter speed when the same aperture was selected. The Tamron histogram and exposure was the closest to the DA16-45, but was ever so slightly overexposed as one of the channels hit the right side of the histogram.

What metering method did you use? Perhaps with a different
metering method, the A28 would have been spot on and the 16-45
would have been under exposed.

Matrix metering for both, so no they were the same and they were taken within seconds of each other . Aperture priority and f8 selected and the camera set the shutter speed. Both matrix metering. I do know how to use a camera.;-)

--
Lance B

 

[John Bean (UK)]

I'm not sure why the difference, but the DA16-45 is spot on. By the
way, all 3 lenses tested, DA16-45, A 28mm f2.8 and the Tamron 28-75
F2.8, used the same shutter speed when the same aperture was
selected. The Tamron histogram and exposure was the closest to the
DA16-45, but was ever so slightly overexposed as one of the
channels hit the right side of the histogram.

The A28 either has a slightly sticky aperture, or maybe a little wear in the stopdown mechanism. All the camera does is move the lever to the position that should represent the aperture selected, but it's down to the mechanics of the lens to make it so. My A50/1.4 is being serviced as I write, after I noticed it was overexposing compared with my FA28-105. Slight oiling of the aperture blades making the diaphragm less "snappy" was enough to produce nearly a full stop of overexposure at f/8 in my case.

--
John Bean

PAW 2005 Week 17:



See: http://waterfoot.smugmug.com/paw