G400 focus issue

Ru85 · Jan 12, 2004

Most people including me have said their G400 fails to lock focus in low and even moderately good light but some have said the images are fine while others are unsatisfied. My question to those that are not satisfied: Are you getting blurred images or just frustrated by the failure of the camera to indicate focus lock?

I took some pictures under controlled conditions to test the ability of the camera to focus in poor light

I took a picture in a dark room with the focus light indicating focus was not locked and another of the same subject with the lights turned up, allowing me to lock focus. Much to my surprise I found negligible difference in sharpness- To prove that the camera lens was not pre-focused before turning out the lights I first locked focus on a subject 18 feet away, then darkened the room, released the focus lock and shot a subject at 4’ - this insured that the lens had to work in the dark almost instantly to autofocus- and it did. (To check if result was due to depth of field I also tried a test shot by locking focus on the subject at 18 ft then shooting the subject at 4 feet without releasing the focus lock. [See page 41 of the manual]- This shot was blurred as expected, proving that the previous result was not due to a large depth of field ) The results are posted here: http://home.comcast.net/~russo857/G400 These were taken at full telephoto in auto mode

I repeated this about a dozen times in different modes with similar results ( had only one problem in angel mode). However very close examination shows that some of the shots taken with the autofocus locked seem slightly sharper than those in which the autofocus cannot lock. Even when the autofocus could not lock, some low light pictures seem a bit sharper than those taken in total darkness. But I find the differences are negligible - usually only visisble when enlarging beyond full size. Also, I took these shots in either total darkness or very low light- conditions which would be a challenge for any camera.

This camera’s autofocus in low light actually exceeded my expectation. But you need to get past the fact that the camera is indicating the focus cannot lock and simply look at the results. My personal opinion is that the focus indicator does not operate properly. Comments?

Jim Cockfield · Jan 12, 2004

Ru85 said:
This camera’s autofocus in low light actually exceeded my
expectation. But you need to get past the fact that the camera is
indicating the focus cannot lock and simply look at the results. My
personal opinion is that the focus indicator does not operate
properly. Comments?

I reached the same conclusion in my testing of the Konica KD-510z (Minolta G500). The focus lock indicator seems to be based on some combination of factors -- mostly weighted towards light levels (which decrease when zoom is used, since about 3 times as much light can reach the sensor at wide angle).

I do believe that there is a QC problem (or design flaw), that can be corrected via a firmware upgrade. Although, I could be wrong, and it's simply a batch of bad components, or a CCD mounting/alignment issue causing some cameras to be worse than others.

My theory is that the camera's designers assumed that the CCD Sensitivity (or output from the supporting chip set), would always be at a constant level, and there is signficant variation from the CCD output (or output from the supporting chip set), causing many cameras to perform worse than expected, with other cameras performing fine.

See both my Part 1 and Part 2 of tests done with my KD-510z (G500) for an explanation of my theory:

http://forums.dpreview.com/forums/read.asp?forum=1024&message=7231540

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Jim Cockfield · Jan 12, 2004

Your G400 uses a different CCD (smaller 1/2.5" CCD, compared to the G500's larger 1/1.8" CCD). The G500 is using a Sony CCD, and I believe that the G400 is using a Panasonic CCD (although I have no way to confirm this for sure).

However, I think the issues regarding both models are probably the same (QC problem with different signal levels coming from the CCD or supporting Chip set, that the camera's designers did not anticipate or correct for).

If I'm right, then a firmware upgrade should be able to correct the problems (by automatically varying the signal strength from the CCD, without user intervention, to achieve the ability for the cameras to better "see" the contrast between pixels when focusing).

I think that's also why we're seeing such a big difference between the reviews.

For example: Steves-Digicams noted suprisingly good low light focus with the G500, but noted poor indoor focusing from the G400. Dcresource.com noted the opposite with the G500 they tested (indoor focus was "not great").

User experiences have also been varied (some users like me, seem to have cameras that focus great -- others do not).

Again, I could be wrong, but the evidence seems to support my theories (focus lock indicator is not accurate, and seems to be weighted on other factors, with EV Compensation definitely impacting REAL focus accuracy.

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Ru85 · Jan 12, 2004

Jim Cockfield said:
If I'm right, then a firmware upgrade should be able to correct the
problems (by automatically varying the signal strength from the
CCD, without user intervention, to achieve the ability for the
cameras to better "see" the contrast between pixels when focusing).

I suspect you are correct. Unfortunately Minolta seems to take little interest - Tried to email them but just got a form reply back stating they have no info available on the subject- Have heard similar compliants from others.
Thanks
Paul

TechniKal · Jan 12, 2004

Ru85 said:
Are you getting blurred images or just
frustrated by the failure of the camera to indicate focus lock?

Both. It's frustrating for the camera not to lock when it should, and shooting without lock results in less crisp focus than shooting with lock in my testing.

Ru85 said:
This camera’s autofocus in low light actually exceeded my
expectation. But you need to get past the fact that the camera is
indicating the focus cannot lock and simply look at the results. My
personal opinion is that the focus indicator does not operate
properly. Comments?

As far as I can tell, the camera is -not- autofocusing when there is no focus lock. My testing seem to indicate that failure to get focus lock results in the camera setting focus to some fixed focus zone. It may vary zones based on lighting or flash setting or zoom or whatever - but it is not autofocusing in the sense that it adjusting focused based on the clarity of or distance to the subject - which is what Autofocus is supposed to do.

Essentially, the camera 'focuses' well because the subjects are typically within its default DOF - analogus to saying that stopped clock tells good time because its accurate when the time of day happens to line up with the stopped time. To test for DOP impact, shoot some macro shots with varying levels of light. Macro mode will reduce the DOP to the point that accurate focus really matters.

Minolta could tell us exactly what the camera is doing with a simple response to multiple emails concerning this issue, but they have never responded. We should send them a bill for QA testing services - they obviously aren't' doing much of it on their own. We should send them a bill for Tech Support services, too - especially Jim C. It's a shame that the most responsive and useful tech support for Minolta comes from one of its customers.

I also do not believe that the focus lock indicator is defective. It may occasionally give a lock indicator when it's not in focus - especially in some of the scene modes, but I believe the lack of lock is a true indication that the camera can not properly focus, and thus goes to some pre-determined default fixed focus zone. There's documentation to this for other Minolta cameras on Minolta's website.

I stand by my asseration that the G400 performs very poorly compared to other cameras in its class in less than optimal lighting. Every other small camera I've tried has locked focus more often and produced consistently more accurately focused photos in typical indoor lighting shots. If you want a small, fast camera for outdoor shots, the G400 is a good choice. If you plan to shoot indoors in normal lighting, the G400 will disappoint compared to other cameras in its class.

Jim Cockfield · Jan 12, 2004

Ru85 said:
I suspect you are correct. Unfortunately Minolta seems to take
little interest - Tried to email them but just got a form reply
back stating they have no info available on the subject- Have heard
similar compliants from others.
Thanks
Paul

Well, bear in mind that these are Konica engineered cameras. I suspect that the merger is having a huge impact on their responsiveness, too.

They are no doubt trying to consolidate resources between the two companies, which may make fixes to known issues take longer than they have in the past (as well as getting these issues to the "right people" in Japan, where the cameras are designed).

The good news is that both Konica and Minolta both have great reputations for firmware upgrades to correct known issues.

Konica-Japan has released firmware upgrades for the Konica KD-310z, KD-400z, KD-410z, and KD-500z to improve many aspects of the cameras performance.

So, I'd also expect the same thing for the new Konica KD-510z (Minolta G500) and the Konica KD-420z (Minolta G400) at some point.

Unfortunately, these models (from Konica) are new to the U.S. support staff, so they have no experience with previous Konica Models (and I suspect that the interface between the Konica-Japan engineering/development groups also leaves a lot to be desired -- thanks in no small part to the merger).

In my case, I'm willing to live with the "quirks" found so far. I think the camera's strengths, far outweigh any weaknesses (at least in my particular camera).

Hopefully, they will eventually "sort out" the merger issues, and improve these models via firmware upgrades, just as they have for previous Konica models. I see no reason why they would not.

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Jim Cockfield · Jan 12, 2004

My tests with my KD-510z indicate that this is simply not the case. I have tested this at length (see my test results part 1 and 2, that I pointed to in my previous response in this thread).

The focus lock indicator seems to be weighted on available light, not true focus lock. This can easily be proven (as I pointed out in my test results).

You can vary EV and get a lock with my camera (yet the photos will be out of focus). Also, you can test easily test this theory by focusing on a "borderline subject" where you get a focus lock at wide angle, but not at full zoom. Then, compare the photos (for actual focus), with fixed focus choices (1m, 2m, 4m). The photos without a lock in my camera using zoom are sharp, where the fixed focus choices are not.

In virtual darkness, this changes (the camera does appear to default to a fixed focus point then, about mid way through the flash range). But, it does not, when any reasonable light and contrast is available, despite the lack of a focus lock indicator. Also, you can easily verify that EV does impact focus accuracy (as well as the lock indicator).

Again, my theory is that the designers assumed that a constant signal level would always be present from the CCD (or supporting chip set), and did not design a way to vary this level so that the camera could better "see" the contrast between pixels (as you can do by varying EV level, which appears to increase or attenuate the signal being used by the camera to focus).

Read my test results part 1 and part 2 for the behavior I noted in my testing of the KD-510z, starting with this post:

http://forums.dpreview.com/forums/read.asp?forum=1024&message=7231540

Could my theory be wrong? Yes, but the evidence I'm getting with my camera supports this theory. For all I know, it's simply a batch of bad components, or a CCD mounting issue causing the variation we're seeing between cameras.

In any event, my tests (with my camera) show that it's not defaulting to a fixed focus point with light present, and that varying EV does impact both focus lock and focus accuracy (based on looking at the actual photos).

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Jim Cockfield · Jan 12, 2004

BTW, I saw one A1 user posting that he had to get his 3rd camera, before getting one that focused accurately. So, I suspect that the same problem could be happening with other cameras using a Contrast Detect Focus system, too (only it's more obvious in cameras using a smaller CCD, since it would be harder for a camera with smaller sensor to see the contrast between pixels to focus).

Jim Cockfield said:
My tests with my KD-510z indicate that this is simply not the case.
I have tested this at length (see my test results part 1 and 2,
that I pointed to in my previous response in this thread).

The focus lock indicator seems to be weighted on available light,
not true focus lock. This can easily be proven (as I pointed out
in my test results).

You can vary EV and get a lock with my camera (yet the photos will
be out of focus). Also, you can test easily test this theory by
focusing on a "borderline subject" where you get a focus lock at
wide angle, but not at full zoom. Then, compare the photos (for
actual focus), with fixed focus choices (1m, 2m, 4m). The photos
without a lock in my camera using zoom are sharp, where the fixed
focus choices are not.

In virtual darkness, this changes (the camera does appear to
default to a fixed focus point then, about mid way through the
flash range). But, it does not, when any reasonable light and
contrast is available, despite the lack of a focus lock indicator.
Also, you can easily verify that EV does impact focus accuracy (as
well as the lock indicator).

Again, my theory is that the designers assumed that a constant
signal level would always be present from the CCD (or supporting
chip set), and did not design a way to vary this level so that the
camera could better "see" the contrast between pixels (as you can
do by varying EV level, which appears to increase or attenuate the
signal being used by the camera to focus).

Read my test results part 1 and part 2 for the behavior I noted in
my testing of the KD-510z, starting with this post:

http://forums.dpreview.com/forums/read.asp?forum=1024&message=7231540

Could my theory be wrong? Yes, but the evidence I'm getting with
my camera supports this theory. For all I know, it's simply a
batch of bad components, or a CCD mounting issue causing the
variation we're seeing between cameras.

In any event, my tests (with my camera) show that it's not
defaulting to a fixed focus point with light present, and that
varying EV does impact both focus lock and focus accuracy (based on
looking at the actual photos).

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Ru85 · Jan 12, 2004

TechniKal said:
It may vary zones based on lighting or flash setting or zoom
or whatever - but it is not autofocusing in the sense that it
adjusting focused based on the clarity of or distance to the
subject - which is what Autofocus is supposed to do.

It's possible that it resets to a zone . In my test I focused on an object definitely outside the DOF of the subject (18') , turned out the lights, then released the focus lock and shot the subject (4'). The lens would have to autofocus instantly in almost total darkness to account for the results I got. I have owned dozens of cameras in my life ( although this is only my second digital.) I don't know of any camera that could do that, although I am not familiar with what has been produced in recent years.

TechniKal said:
Minolta could tell us exactly what the camera is doing with a
simple response to multiple emails concerning this issue, but they
have never responded. We should send them a bill for QA testing
services - they obviously aren't' doing much of it on their own.
We should send them a bill for Tech Support services, too -
especially Jim C. It's a shame that the most responsive and useful
tech support for Minolta comes from one of its customers.

It is frustrating to get a canned response like "no info is available on this subject" . Even if they needed time to study the issue as Jim suggests they could communicate better
Thanks

Ru85 · Jan 12, 2004

Hope you are correct- will give them the benefit of the doubt for now

Jim Cockfield said:
Ru85 said:

I suspect you are correct. Unfortunately Minolta seems to take
little interest - Tried to email them but just got a form reply
back stating they have no info available on the subject- Have heard
similar compliants from others.
Thanks
Paul

Click to expand...

Well, bear in mind that these are Konica engineered cameras. I
suspect that the merger is having a huge impact on their
responsiveness, too.

They are no doubt trying to consolidate resources between the two
companies, which may make fixes to known issues take longer than
they have in the past (as well as getting these issues to the
"right people" in Japan, where the cameras are designed).

The good news is that both Konica and Minolta both have great
reputations for firmware upgrades to correct known issues.

Konica-Japan has released firmware upgrades for the Konica KD-310z,
KD-400z, KD-410z, and KD-500z to improve many aspects of the
cameras performance.

So, I'd also expect the same thing for the new Konica KD-510z
(Minolta G500) and the Konica KD-420z (Minolta G400) at some point.

Unfortunately, these models (from Konica) are new to the U.S.
support staff, so they have no experience with previous Konica
Models (and I suspect that the interface between the Konica-Japan
engineering/development groups also leaves a lot to be desired --
thanks in no small part to the merger).

In my case, I'm willing to live with the "quirks" found so far. I
think the camera's strengths, far outweigh any weaknesses (at least
in my particular camera).

Hopefully, they will eventually "sort out" the merger issues, and
improve these models via firmware upgrades, just as they have for
previous Konica models. I see no reason why they would not.

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Jim Cockfield · Jan 12, 2004

With lights out (virtualy darkness), my G500 appears to default to a fixed focus distance, probably about mid way through the flash range.

But, with reasonable light (one light in a room), it's focus does not default to this distance.

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Jim Cockfield · Jan 12, 2004

Ru85 said:
Hope you are correct- will give them the benefit of the doubt for now

Even if they never fixed the "quirks", I still wouldn't trade my KD-510z (a.k.a., Minolta G500) for any other pocketable camera.

Ru85 said:
Jim Cockfield said:

Ru85 said:

I suspect you are correct. Unfortunately Minolta seems to take
little interest - Tried to email them but just got a form reply
back stating they have no info available on the subject- Have heard
similar compliants from others.
Thanks
Paul

Click to expand...

Well, bear in mind that these are Konica engineered cameras. I
suspect that the merger is having a huge impact on their
responsiveness, too.

They are no doubt trying to consolidate resources between the two
companies, which may make fixes to known issues take longer than
they have in the past (as well as getting these issues to the
"right people" in Japan, where the cameras are designed).

The good news is that both Konica and Minolta both have great
reputations for firmware upgrades to correct known issues.

Konica-Japan has released firmware upgrades for the Konica KD-310z,
KD-400z, KD-410z, and KD-500z to improve many aspects of the
cameras performance.

So, I'd also expect the same thing for the new Konica KD-510z
(Minolta G500) and the Konica KD-420z (Minolta G400) at some point.

Unfortunately, these models (from Konica) are new to the U.S.
support staff, so they have no experience with previous Konica
Models (and I suspect that the interface between the Konica-Japan
engineering/development groups also leaves a lot to be desired --
thanks in no small part to the merger).

In my case, I'm willing to live with the "quirks" found so far. I
think the camera's strengths, far outweigh any weaknesses (at least
in my particular camera).

Hopefully, they will eventually "sort out" the merger issues, and
improve these models via firmware upgrades, just as they have for
previous Konica models. I see no reason why they would not.

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Click to expand...

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Ru85 · Jan 12, 2004

I have not done the testing you have but have noticed in some cases, shooting same subject, that low light pictures are very slightly better than no light pictures. In the cases that I have not seen a difference, the zone + DOF might render the difference insignificant. I tested VERY low light(just enough to make out the subject in the LCD) vs almost total darkness

Jim Cockfield said:
With lights out (virtualy darkness), my G500 appears to default to
a fixed focus distance, probably about mid way through the flash
range.

But, with reasonable light (one light in a room), it's focus does
not default to this distance.

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

Jim Cockfield · Jan 12, 2004

Ru85 said:
I have not done the testing you have but have noticed in some
cases, shooting same subject, that low light pictures are very
slightly better than no light pictures. In the cases that I have
not seen a difference, the zone + DOF might render the difference
insignificant. I tested VERY low light(just enough to make out the
subject in the LCD) vs almost total darkness

I suspect it works the same way as my KD-510z (G500). You can see the G500 "gain up" the image momentarily in the LCD display (no doubt so that it can "see" the contrast between pixels better to focus). It then uses this "gained up" image to set focus, regardless of whether or not it gives you a lock.

With no light, I get totally different results (with the camera defaulting to a range that is relatively close to the camera).

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

photoshooter · Jan 12, 2004

Jim Cockfield said:
Ru85 said:

I have not done the testing you have but have noticed in some
cases, shooting same subject, that low light pictures are very
slightly better than no light pictures. In the cases that I have
not seen a difference, the zone + DOF might render the difference
insignificant. I tested VERY low light(just enough to make out the
subject in the LCD) vs almost total darkness

Click to expand...

I suspect it works the same way as my KD-510z (G500). You can see
the G500 "gain up" the image momentarily in the LCD display (no
doubt so that it can "see" the contrast between pixels better to
focus). It then uses this "gained up" image to set focus,
regardless of whether or not it gives you a lock.

With no light, I get totally different results (with the camera
defaulting to a range that is relatively close to the camera).

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

This is crazy.. everyone is trying to figure out the focus... yet are telling people to buy this camera!!!

TechniKal · Jan 12, 2004

I finally got a response from emails to Minolta about this issue. The response was to return the camera to my vendor if I'd bought it recently, or to send it to them for repair/evaluation. No details of whether what I am experiencing is normal for this camera or not.

fannta · Jan 13, 2004

Jim Cockfield said:
I suspect it works the same way as my KD-510z (G500). You can see
the G500 "gain up" the image momentarily in the LCD display (no
doubt so that it can "see" the contrast between pixels better to
focus). It then uses this "gained up" image to set focus,
regardless of whether or not it gives you a lock.

With no light, I get totally different results (with the camera
defaulting to a range that is relatively close to the camera).

--
JimC
------
http://www.pbase.com/jcockfield/konica_kd510z

As far as i have seen with my KD-420Z it does not "gain up" the image when half-pressing the shutter button , the only thing that happens , is that you can see the colours "changing" for a short time while the correct WB is detected/applied .

Actually the "gaining up" and "down" happens all the time when paning around the cam , if pointing at something dark ist gains up , pointing at things with a lot of light the lcd gets darker to compensate for it all this happens continously without having to depress the shutter button .

Ru85 · Jan 13, 2004

photoshooter said:
photoshooter said:

This is crazy.. everyone is trying to figure out the focus... yet are telling people to buy this camera!!!

Click to expand...

I think I am the only one here who has done that - If you read the beginning of this thread & check the pictures I posted, you will see that I am getting good results from this camera. I did not recommend it until I tested it carefully & determined that mine is focusing fine and the issue seems to be only with the focusing indicator not the focus.

I bought this camera before any reviews were posted. If I had seen the complaints about the focus lock in low light I probably would not have bought it. But it turns out - at least on mine and some others who have reviewed this camera- that even though the focus indicator is saying focus is not locked, the images are in focus. For me that changes this issue from a major problem to a minor annoyance. Others have had a different experience or different opinion.

f1champ · Jan 13, 2004

TechniKal said:
I finally got a response from emails to Minolta about this issue.
The response was to return the camera to my vendor if I'd bought it
recently, or to send it to them for repair/evaluation. No details
of whether what I am experiencing is normal for this camera or not.

I got the same thing within a few days from Minolta. I didn't return since I figured all of them are the same.

TechniKal · Jan 13, 2004

f1champ said:
I got the same thing within a few days from Minolta. I didn't
return since I figured all of them are the same.

I'm thinking pretty much the same thing - it seems that what I'm experiencing is normal with the G400. I have forwarded on Minolta's response to Amazon, though. If Amazon will accept the camera back for a full credit, I'll likely go that route (doubtful, given that the camera is well used). Otherwise, I'll keep it, be happy with it's outdoor performance and hope that Minolta fixes the bad indoor performance sometime in the future.

G400 focus issue

Member

Jim Cockfield

Guest

Jim Cockfield

Guest

Member

Active member

Jim Cockfield

Guest

Jim Cockfield

Guest

Jim Cockfield

Guest

Member

Member

Jim Cockfield

Guest

Jim Cockfield

Guest

Member

Jim Cockfield

Guest

Well-known member

Active member

Well-known member

Member

Member

Active member

Similar threads

Keyboard shortcuts