CP 950 Frequently asked questions, and issues. Last update November 10, 1999
Here are some interesting issues regarding the CP950. This isn't a real FAQ, which would have all the details about operating the camera, just what I consider to be the interesting bits. This is all unofficial. I have no connection with Nikon. These are my personal opinions. I have tried to personally confirm each of these items. This has ended up being a list of problems and interesting quirks, mostly. The list of what is "right" with the CP950 would be MUCH larger. I would recommend the CP950 to anyone. Most people would never even notice most of these things, I present them here for your entertainment. Some of this material is now dated (04/08/2000)
CP950 Issues, Features, and FAQs
I have changed this section to reflect changes in the new 1.2 firmware.
Automatic number sequence fails. (Fixed in firmware 1.2)
The camera had the sequential numbering bug in version 1.0 and 1.1. This was listed as having been fixed in the 1.1 firmware, but it was not. It has been fixed in 1.2 by the addition of a SEQ.XFER setting in A-Rec mode
The symptom was, that after clearing off your flash card, (any way that you want to), the first pictures that you take in A-REC mode will be numbered 0001, 0002, etc. Once you take a picture in M-REC mode, the sequential numbering will be correct, even for A-REC mode. As a work around, people have been taking a small low resolution shot in M-REC and leaving it on the camera. (protect the shot).
The "out of range" flashing shutter speed or aperture no longer works on the top LCD with firmware 1.1 or 1.2 on some cameras.
In firmware 1.0, if you set shutter or aperture priority mode, and you were out of range (too dark), the top LCD would flash with the aperture or shutter speed. This no longer works (it does flash on the color LCD screen). Strangely, this does not affect all cameras! One person has reported that his still flashes out of range with both 1.1 and 1.2. Mine does not work with either. This is bad because you really can't operate the camera in priority modes with the LCD monitor off without that indication.
Fisheye lens is out of focus with version 1.1 firmware.
Or so it has been reported. I don't have the Nikon fisheye lens to test this with. Supposed to have been OK with version 1.0 firmware. Still a problem with 1.2. Nikon tech says that this is a focus adjustment problem with individual cameras, and can be adjusted at the repair center. This really bothers people who need to use the fisheye to make panoramic shots. I do not know what percentage of cameras are actually bothered by this, the fisheye lens is not a very popular accessory, so the reports are scattered.
The camera zooms in all the way when turned on.
Every time you turn the camera on, or switch to play mode and back to record, or let the camera go to sleep and wake it again by pressing the release, the camera lens will return to the fully zoomed position. You will have to re-set it ever time. It takes longer to fix it every time than if the camera returned by itself. This behaviour is unchanged in firmware 1.2.
(Note, the camera is not set QUITE at full zoom. It is about 95% of the full zoom setting)
I really hate this myself. There are probably a few people that like this I suppose, but this has to be the top complaint about the camera.
Speculation: One theory is that the camera likes to leave the lens in a certain position for safe storage, or to get leverage when starting, or to calibrate the encoder for the lens, and that the zoomed position is it. That is why when you turn the camera off, it makes an extra effort to "park" the lens before actually turning off. However, the behavior when you let the camera go to sleep does not correlate with this. When the camera sleeps, it just turns off. When you wake it up, it makes an extra effort to zoom the lens in all the way (maximum zoom) before giving you control. This isn't consistant with the "safe parking" theory. It might be consistant with the "calibrate the lens theory", but there would be no reason that the lens could not then move to any position that you want after that.
Why doesn't the camera zoom back in when you turn it on? Well, maybe the designers said "The reviewers will be timing the turn on time" "Let's make it as fast as possible by not taking the two seconds to return the lens". Unfortunately, this means that you have to do it yourself; every time, which is actually slower.. I wish that this was optional
My Alternate theory. Just a theory. Maybe the camera uses an optical encoder to sense movement of the zoom lens, instead of a more expensive stepper motor. Because the camera just "counts pulses" to track where the lens is, it needs to return the lens to a known position every time the power cycles to start the count over. Maybe they just crank the lens to maximum zoom, or to a stop or switch to do this. Why not just store the pulse count between power cycles? Well, every now and then there might be a pulse missed, or the lens might move with the power off. Then the camera could no longer be sure that it knew where the lens was. Also maybe there is no end of travel sensor. The only way that they have is to start the lens from one of the extremes and count from there.
If the alternate theory is correct, it might be harder to fix in firmware. The camera would still have to crank the lens out to the telephoto position, but there would be no reason why it could not then return to the 50mm position, say. This would take longer though! Everyone who wanted a fast turn on got their wish.
Sort of Workaround:
Providing that you don't need to use the flash, if you go to the lens selection and pick the wide angle adapter, the camera will start out zoomed all the way to wide angle, which is better than starting at full zoom if you ask me.
Possible solution suggestions:
1) Let the flash turn on in wide angle mode.
2) Return the lens to the position you left it. (my choice)
3) Provide a lens choice on the lens menu or controls menu for "start at 50mm"
4) Provide a lens choice with several settings! 38, 50,70, 110
5) Just start at 50mm, who cares how long it takes!!!! It will actually take less time than having to do the full zoom.
When taking a low light shot, the LCD screen gets darker after focusing. (Improved in firmware 1.1) (Really improved in 1.2)
This has really been improved in version 1.2. Read about it on the firmware 1.2 page.
In firmware version 1.1, the camera LCD does not dim as much as it did previously, allowing you to frame your subject. I would now call it "good enough". There is still some dimming in very low light, but you can now see enough to frame your subject. I also note that the camera exposure estimate does not change as it used to after you half press to pre-focus.
Here is the explanation of what is going on, but it is less interesting now that Nikon has improved this.
This is a complicated thing. First, to see the effect, point the camera at a low light area. The area will be visible in the LCD until you press the release 1/2 way to focus. After the camera grabs a focus, the LCD will darken; maybe even so much that you can't see anything any more!
At first I thought that what was happening was a "depth of field preview" where the camera was stopping down to the f-stop that it was actually going to use for the shot so that you could see the depth of field on the screen. However, you can prove that this isn't the case with the following test.
Force the flash off. Compose a shot in low light that will require full aperture. (say f2.8). Now press 1/2 way to focus. It still gets darker on the LCD even though the lens is going to be used all the way open.
What seems to be happening acutally is that the camera increases the gain on the CCD during focusing, maybe even beyond what is allowed with the sensitivity adjustment. I can think of several reasons: 1) It lets you see better in low light 2) it allows a fast frame rate on the LCD, people complain about cameras with low LCD frame rates 3) it gives the camera more to work with for it's focusing procedure. Once the focus is aquired, the camera returns to normal gain (except that you can't always see the shot any more). It may do this for another "speed up" reason; to take the shot as quickly as possible after the release is pressed; another thing that reviewers time. It could be that it takes time for a gain adjustment to settle in, and they are getting ready for the shot.
The main problem with this is that after the gain is reduced when you half press to focus, and the focus should be locked, it then the camera seems to try to focus again with the gain turned down when you press the release. Since the gain is turned down, it usually spoils the focus on your shot. It really should lock the focus after the half press.
(This basic scenario has now been confirmed on the Nikon support BBS)
The camera will not always retain the sensitivity adjustment (Fixed in firmware 1.1)
INF picture bug. (Fixed in firmware 1.1)
Once you take a shot at the INFinity setting, all further pictures will have their EXIF header labled as INF focus until you turn the camera off and on. You can see this yourself; take normal picture; switch to play mode and look at the EXIF information (spin the wheel) to see that it says FOCUS:AF. Now take an INFinity lock picture. Switch to play, it now says FOCUS:INF. That is good. Now turn INF off, and take any shots that you want, they will all still say INF.
Flash mode changes by itself.
There are two ways that this can still happen in firmware 1.2 involving A-Rec mode. If this happens to you, read about it on my firmware 1.2 page.
Low light focusing
I will have more commentary later on this issue also. In brief, my inital impression is that the low light focusing may have a few "tricks", and that it is roughly as good as on the CP900 (but no better), and not as good as some other cameras. The low light focusing is much better than my PDR-M1 camera. My impression is that the low light focusing is actually quite good, but you may have to make several attempts to grab a focus. I think that the comments about poor focusing are due to the following factors.
* The camera can focus in much lower light than the old CP900, and the screen is much brighter, so there is the expectation that the camera should be able to focus. In side by side comparisons, the CP950 could focus in situations where the CP900 had an amost black screen.
* The camera will select low shutter speeds that cannot really be hand held. The old rule of thumb is 1/focal lenght, so no longer than 1/38 second for wide angle, 1/50 sec for 50mm, etc. Trying to hand hold a 1/15 or 1/4 second exposure almost always results in blur, which can look like bad focus. Also note below, the exposure shown on the screen is just approximate, it may actually be slower than what is shown! I really think that the fact that the low light shots that most people take are hand held accounts for a lot of the reports of poor focus in low light.
* The camera needs some contrast to focus. You can get it to fail by pointing it at a white wall in sunlight.
Noise on low light shots, and what to do about it.
I have been taking quite a few. The low light sensitivity is great. Until then, know that some amount of noise and "hot" pixels are normal on shots over 1 second. See my section on CP950 astro shots for pictures of stars and talk about dark frames. It is "normal" to see some bright pixels on long exposures with the camera. If your camera does not show any obvious bright pixels at default ISO and 1 or 2 second exposure with the lens cap on in the dark, then I think that the CCD is acceptable. 8 second shots at ISO 320 can look like a galaxy of stars.
No shutter priority with higher ISO modes.
This is not a bug, but rather a limitation of the camera. You can't use the shutter priority with the higher ISO modes. I initially guessed that this was done because of the amount of noise that would result from 8 second exposures at the higher ISOs, however since there is another way to get 8 second exposures at high ISO, and since they look OK, I'm not sure why the limitation is there. The highest exposure allowed in the Program mode is 1 second in all ISO settings. You can do a lot with 1 second at ISO 320. HOWEVER. You can get around this pretty much at will by picking the Shutter Priority mode at High ISO, and turning off the flash. The camera will pick up to an 8 second exposure. On my camera, the exposure will sometimes keep changing between 4 and 8 seconds (noise on the CCD?), but if you press while the 8 second exposure is showing, you will get 8 seconds.
There may be a reason for this, although this is unofficial. The camera only has three f stop settings. To control the exposure more finely, the camera usually adjusts the shutter speed. In shutter priority mode, it can't so this, so it adjusts the ISO instead to get a good exposure. This is why it will not allow you to adjust the ISO yourself in shutter priority mode. There does not seem to be any indication of the adjusted ISO in the EXIF header in the images taken though.
Color issues and the flash.
The camera does seem to have some issues with the flash. The main problem is that the white balance IS NOT SET TO FLASH in M-REc mode. You need to set it yourself!
1) If you mix "fill in" flash with other types of lighting, the white balance can be off, and your shot might have a "tint". I think that this is a very difficult task for the camera. I have one shot in particular in mind where there was low natural light coming in through a window, a flourescent light was on, and the camera also decided to flash. The entire picture has an olive green tint. It might be expecting the camera to do to much to have good white balance in that situation.
2) If you use the flash for "fill in" by forcing it on, the camera does not compensate the exposure for the fact that the flash is on. This generally results in overexposure. I would say that at this point not to use the fill in flash if the subject is closer than 6 feet. It might be possible to find an EV compensation setting that would work, but it will be hard to judge what setting to use in each case. This and the mixed light problem above make it difficult to get good fill in flash action. Other cameras, such as my Toshiba PDR-M1, do a very good job of fill in flash and color balance.
Lockups. (Fixed in firmware 1.1)
This section is obsolete now. I keep it here for anyone still running 1.0 firmware.
One of the improvements of the new firmware is supposed to be fewer lockups. I have not used the camera enough after upgrading to be able to comment yet (07/22/1999) The following applies to my camera with the 1.0 firmware.
I've had the 950 lock up on me twice now, after about 550 and then 750 shots. The first time, I was taking pictures of the moon in MREC mode, using shutter priority. The first symptom was that the camera stopped responding to the shutter changes. It was intermittent. Then the camera stopped responding entirely, and the LCD monitor turned off. I turned the camera off with the switch, but it would not turn off, the LCD on top still showed record mode. I let it sit for 5 minutes in the Off postion, but it would not turn off. I had to remove and replace the batteries, and it snapped out of it. I've taken about 200 shots since then as of this writing, and it has been OK. The batteries were not showing low, by the way. The second lock up was in a similar situation in shutter priority mode after a long session trying out features. Flash, I have now had a third lockup, at about exposure count 900. In the last case, I was just carrying the camera around in the "on" position, after having taken some shots. I am sure that these lockups that I have had are not low battery problems.
There seem to be at least three types of lockups, and there might actually be more. One type, is related to the serial transfer, and involves unplugging the cable at the "Wrong" time. Another type is that the camera will go to "sleep" normally, but then can't be woken up again. I have had this one once. A third type is that the camera will lock up during use, usually after heavy use. The LCD on top will remain on, and the camera can't be turned off.. There have been reports of the color LCD screen being stuck on too. All of the lockups can be cleared by opening up the battery door for a second.
My personal feeling is that these are software related; if I had to guess I would say that there is a memory leak in the operating system. Because of this, your own operating habits might affect how often you see a lockup; if you tend to leave the camera on a lot, and shoot a lot of pictures, then you will see more. If you turn the camera on and off a lot, so that the software can reset, you might never see them. Lately, I have been forcing myself to turn the camera off more instead of letting it go to sleep, and I have not seen any lockups since then.
Other posters on the various forums have also suggested that the lockups might just be due to low batteries, or to the brand of flash card that people are using. This does not match my experience, but there is certainly room for low battery and bad flash card lockups in the spectrum of lockups. Only time will tell. If a future firmware upgrade fixes most of them, then they will have been proved to be software.
Here is one way to get the camera to lock up, if you want to try it. Sent in by Mareck Chudzia, first discovered and posted on the Nikon BBS by Doug Larrick.
(This lockup procedure no longer works with firmware version 1.1)
LCD on "review only", quality on "Normal", 1600x1200, no other fancy features selected.
1. Camera off
2. Main switch to "M"
3. Push flash button once to turn off flash
4. Push Macro button twice to go into macro mode
5. Push the Monitor button to turn on the monitor
6. Compose shot, let it focus, take picture
7. While "delete/pause" screen is up, push shutter halfway Note that at this point, the monitor turns off but cannot be turned back on (you are doomed now)
8. Turn the main switch to off". Lens moves (parks?) and camera is now locked up solid. Must open battery cover to reset.
In my testing of this, I found that my camera would lock up ever time with Mareck's procedure. I could also get it to lock up sometimes by taking pictures with the hourglass on in the fast review mode. You can see how this and other lockups, which are obviously software, cause people to think that there is something wrong with their camera.
Camera stops responding to changes in Shutter Priority mode (Fixed in 1.1).
Camera stops being able to focus (Fixed in 1.1).
Distances shown in Manual Focus mode are not always accurate.
This is still true in version 1.1 and 1.2. See my page on this here for more detail.
This is true for the closer settings. In fact, I suspect that they are only accurate at about 50mm zoom (unfortnuately, there is no way to select 50mm zoom except by guessing, so...). For example, at wide angle, with the manual focus set to 0.33 feet, the camera focus is found at about 2.5 inches, or about 0.2 feet. At maximum zoom, the focus is at about 7 inches, or 0.58 feet. (Thanks to Dan Lauring for noticing this effect). This applies especially if you have the TC-E2 telephoto lens attached; then infinity focus is about at the 10 foot setting. I was really hoping that this would change in the new firmware.
The worst ramification of this is that the INFinity settings are usually not sharp. This bites new users over and over as they try to take pictures with telephoto at infinity.
Green blotch in self portrait mode. (Fixed in 1.1)
Mystery little white pixel on LCD monitor.
This isn't a problem, but is weird. Try this. Point the camera at a dark object (so you can see the mystery pixel, which will be white). Watch the LCD when you half press to focus. As you first start to press, a single white pixel. will illuminate briefly on the LCD monitor. The pixel is located on the right side, 8mm from the top, and 3mm from the right side. It must either mean something, or is a "debugging" feature left in by accident. The Mystery Pixel is still present in firmware version 1.1 and 1.2
I make changes, but the exposure does not change.
You take the camera and point it at an indoor area. You look at the exposure. It says 1/30 at say second (say at f2.6). You increase the ISO (say from 100 to 200), or turn on more room lights. The exposure stays the same at 1/30. Why? It is because the light is low enough that the camera is going to use the flash. It always uses 1/30 for the flash. Until you exceed the light level where it will no longer want to use the flash, it will always use the same exposure. This isn't really a bug so much as operator; error; a common new user question
Best Shot Selector does not allow 10 shots.
In the "pocket guide" and on the Nikon web site in the CP950 specifications, it states that the BSS feature will take pictures as long as you hold the button down, up to 10 shots. I never get 10 shots. Here are
the results of several
This was a problem just today 04/29/1999 because I was trying to take some pear blossoms blowing in the wind, and the camera would only take 4 or 5 shots per BSS usage, so I had to repeat each one several times.
BSS does not even seem to work
There is also a discussion of whether the best shot mode even works at all. I think that it does, but not in the way you might expect. Here is my reasoning....
Easter Egg / firmware version display
Hold down the menu button while you turn the camera on. You will get a screen showing the firmware revision. If you turn on the camera this way directly into PLAY mode, the camera will lock up, however, and cannot be turned off, so don't do that. The camera will still lock up in Play in firmware version 1.1
This is total speculation on my part. This is not official infomation.
First, lets go over the evidence...
* The camera obviously does not have a 48MB buffer as was reported for the pre-production models. This has been officialy confirmed by Mike Rubin from Nikon. The actual size is not known.
* In continuous mode, using FINE, you can only take 5 shots. (roughly 4 800K shots, and 1 1.2MB shot, see below)
* There is a weird bug in continuous FINE mode where the last shot taken is much bigger than the rest.
* In BSS mode, using FINE, you will never be able to take the 10 shots that are mentioned in the manual. You will get 4 to 9 shots, with about 6 being the average.
* The camera has to have room to store one TIFF shot, so I think that this sets the minimum size for the buffer.
* The camera appears to be able to compress a shot to JPEG in a fraction of a second.
* The camera has a very fast focus routine that works off the image. Consider how many focus decisions it must have to make to focus in less than a second, when there are over 4000 focus positions (obviously it does not stop at all 4000, but it probably takes a reading, jumps to a new postion, takes a reading, guesses at the proper focus, jumps there, takes a reading, moves a little bit in one direction, is that better or worse.... etc. There must be a very fast "is this sharper than this" routine.
* RAM is expensive, there is no sense using more than needed.
Here is my theory.
First, the BSS. I think that the BSS actually only needs to store two images, the "best so far" and the current image. How is compares these images is unknown, maybe it uses its fast focus routine or maybe it compresses them to JPEG and takes the largest one. If you think about it though, it only needs two images. Once it has decided that one image is worse than another, it can discard it. Why it takes fewer than 10 shots in FINE mode I don't know. Maybe there is a time limit.
Now on to the buffer size.
For continous mode, it has to store all of the images. 5 in fine mode. I think that these are compressed to JPEG as you take them. The buffer must be large enough to hold them all. On my tests, the total size of the images taken was up to 4.6 MB.
The buffer must be large enough to hold at least one TIFF file, 5.6 MB.
These two items set the minimum size, the buffer has to be at least 5.6 MB.
Does the buffer have to be any bigger than the minimum size? Well, it depends on how the BSS mode actually works, and how the images are stored in the camera. and how much "scratch space" is needed. If images are stored as "raw" CCD data, they will take 1.92 MB. If as RGB data, then 1.92 x 3 = 5.76 MB. If JPEG, then up to 900K each.
If the images are stored as JPEG, then the buffer could be just 6MB to hold continous mode shots. If BSS actually stores all intermedate shots (why?) the the buffer might have to be as large as 9x900KB=8.1 MB, but I doubt it (If the buffer were only 6, that might explain why you can't take 9 BSS shots in FINE mode some times).
There is strong evidence that the camera can compress to JPEG very fast. When the camera has to store more that one shot, the shots are probably stored in the buffer as JPEG. BSS probably makes it's selection "on the fly" so that it does not have to store all the shots. A 6 MB buffer would be sufficient to explain all of the observed behaviour. I think that the buffer is only (about) 6MB. (possibly a bit larger if some "scratch space" is needed, say an extra 1MB).
Not really a 3X zoom!
The camera does not appear to actually be a 3X zoom. One of the published numbers must be off; I suspect the high end. Here is the evidence.
1) In the EXIF header, the focal lengths show as 7.2mm for wide, 20.4mm for zoom. This is a ratio of 2.83. Well, that could be wrong.
2) I took a picture of a ruler square on using a tripod. I measured the size of a three inch portion in the center of the image.
Wide angle - 224 pixels, Zoom, 646 pixels. Ratio of 2.88. Maybe I was too close and there was parallax or something.
3) Took a picture of a house from about 100 yards away, measured a garage door at the center.
Wide angle - 495 pixels, Zoom = 173 pixels. Ratio of 2.86.
Conclusion, actually about a 2.87X zoom. Someone somewhere has rounded up.
It this is true, one of the published numbers must be off. A widely quoted number for the CP950 is 38mm to 115mm, which would be 3.02X. Since the wide angle appears to actually be 7.2mm, and the Zoom appears to be 20.4mm, if the wide angle is really 38mm, the camera would actually be 38/7.2x20.4= 107.6 mm on the high end; closer to 110mm.
That is not your actual exposure either!
The exposure shown on the screen is just approximate, apparently rounded to the nearest standard shutter speed. For example, you might take a shot when the screen shows 1/30 sec, but when you examine the picture, you will find it was actually taken 1/24 sec. Take a shot and try it yourself, examine the details on the shot in play mode by spinning the "command" wheel. This can be very noticable at the higher shutter speeds, where for example you might have 1/250 showning on the screen, but get 1/330 sec on the exposure. The exception to this rule is in shutter priority mode, where the camera uses the exact shutter speed that you picked. Apparently it boosts the gain on the camera to compensate for the small number of aperture settings availible.
This is not a bug, if the camera did not do this, your exposure would jump all over like crazy and you could not really read it.
AE lock feature white balance malfunction
After locking the exposure with the first shot with the AE lock on, subsequent shots often have messed up white balance, usually a redish tint on landscapes. NOT fixed in version 1.1 fimware. Appears to be fixed in version 1.2