Originally Posted by Tom Frank

Mine's not nearly as nice.. just 4 in. Celestron reflector. I have viewed Saturn's rings with it, though not nearly as clearly as the pic!.

It doesn't look as good through the eyepiece as it does in that picture. That picture is the result of taking 1200 frames of video and combining them into one picture. The software throws out anything that is not in all the frames, so you get incredible results.

OK..... Back to the camera discussion....

Doug

I've gotten to the point where my final lens tweaking will be done during a bright sunny day. According to the weatherman, we can expect one here sometime about mid-March, so I'm going to post what I've found helpful for me to get that "best" focus. But I'll give fair warning you have to be anal about it to do this!

I think most of the camera videos I've seen posted here show the lens comes fairly well focused from the vendor, so unless you really don't like what you see, my best advice is not to bother with refocusing! I'll skip the part about removing the glue on the lens so it can turn and taping over all exposed circuit board components to avoid shorting and static discharge, and get right to the actual focusing nitty gritty.

I could NOT convince myself I had the best focus by viewing a web cam display on my PC because of too much motion while hand holding the camera, difficulty in turning the lens and not blocking the view at the same time, and no way to really compare one setting with the prior one at both far and near distances other than by memory. But YMMV, and this may be good enough for you. My goal was to optimize distant object focus (for AV purposes), while getting the depth of field as close in as possible before the distant focus deteriorated. And good focus across the whole video frame if possible.

I used my focus as received since it was pretty good and a good starting point, then I marked one of the "indent valleys" in the knurled ring on the lens barrel with paint (so it doesn't scrape off while focusing), with a matching line on the fixed base. There are twelve of these valleys, so they are 30 deg. apart, and I'll call the rotational angle between two valleys in the following procedure as one "notch".


Here's what I did:

• Find a fixed outside scene with both far and near objects for comparison. I like bare tree limbs several hundred feet away as a minimum for distant objects, and just about anything for up close. Also have a spot where you can hold the camera steady and aim the camera the same during test video clips.
• Take a short video, holding the camera steady on a distant object, a mid-range object, and a close-in (e.g. 8-10 ft. away) object. Each time you shoot a clip, save and rename it so you can tell the lens position from the name.
• Rotate the lens one notch clockwise (CW) to move the focal point further away from the camera, and shoot a similar video.
• Repeat the process with the lens rotated one notch counter clockwise (CCW) from the original to move the focal point closer to the camera..
• Compare the CW video with the original, then the CCW locaton with the original. To do this open one rotation video and the original separately with the AviDemux editor, so you have two editor screens arranged side-by-side and can step through the clips frame-by-frame to get still images you can now very easily compare for focus on far, middle and near objects. Do the same comparison of the original position with the opposite lens rotation video. You should now be able to determine which direction of rotation should be fine tuned. If you can't, maybe the lens is focused really good in its original position and you can reposition it there are done!
• If you can see a difference, repeat the process on that rotation side with a smaller rotational increments. I stepped through iterations of 1/4, 1/2, and 3/4 notch positions, comparing each with the prior (I told you I was anal!).
• With this process, you should be able to find two lens positions where the focus appears about the same in both.

I finally found that I could see virtually no significant difference between the 1/4 and 1/2 notch CCW positions from the original location on my camera. It's interesting when you step through frame-by-frame to see how the H.264 compression treats the over all image, with a small area in the video appearing to go into and out of focus and back again from one frame to the next. So it becomes tough to decide which of the two clips is in better focus overall with these small lens rotational increments, but that just means you are pretty much at the sweet spot. I found three positions gave pretty much the same distant focus, with very slight differences in the close focus point. So my final lens resting spot will be based on which of three gives the better close focus. I'll finish that when a sunny days appears. I also saw that good focus in the lower left corner of my videos was almost impossible to achieve, so the cheap lens is showing it's faults there.

Good luck with your focusing efforts!

Originally Posted by Tom Frank

I've gotten to the point where my final lens tweaking will be done during a bright sunny day. According to the weatherman, we can expect one here sometime about mid-March, so I'm going to post what I've found helpful for me to get that "best" focus. But I'll give fair warning you have to be anal about it to do this!

I think most of the camera videos I've seen posted here show the lens comes fairly well focused from the vendor, so unless you really don't like what you see, my best advice is not to bother with refocusing! I'll skip the part about removing the glue on the lens so it can turn and taping over all exposed circuit board components to avoid shorting and static discharge, and get right to the actual focusing nitty gritty.

I could NOT convince myself I had the best focus by viewing a web cam display on my PC because of too much motion while hand holding the camera, difficulty in turning the lens and not blocking the view at the same time, and no way to really compare one setting with the prior one at both far and near distances other than by memory. But YMMV, and this may be good enough for you. My goal was to optimize distant object focus (for AV purposes), while getting the depth of field as close in as possible before the distant focus deteriorated. And good focus across the whole video frame if possible.

I used my focus as received since it was pretty good and a good starting point, then I marked one of the "indent valleys" in the knurled ring on the lens barrel with paint (so it doesn't scrape off while focusing), with a matching line on the fixed base. There are twelve of these valleys, so they are 30 deg. apart, and I'll call the rotational angle between two valleys in the following procedure as one "notch".


Here's what I did:

• Find a fixed outside scene with both far and near objects for comparison. I like bare tree limbs several hundred feet away as a minimum for distant objects, and just about anything for up close. Also have a spot where you can hold the camera steady and aim the camera the same during test video clips.
• Take a short video, holding the camera steady on a distant object, a mid-range object, and a close-in (e.g. 8-10 ft. away) object. Each time you shoot a clip, save and rename it so you can tell the lens position from the name.
• Rotate the lens one notch clockwise (CW) and shoot a similar video
• Repeat the process with the lens rotated one notch counter clockwise (CCW) from the original.
• Compare the CW video with the original, then the CCW locaton with the original. To do this open one rotation video and the original separately with the AviDemux editor, so you have two editor screens arranged side-by-side and can step through the clips frame-by-frame to get still images you can now very easily compare for focus on far, middle and near objects. Do the same comparison of the original position with the opposite lens rotation video. You should now be able to determine which direction of rotation should be fine tuned. If you can't, maybe the lens is focused really good in its original position and you can reposition it there are done!
• If you can see a difference, repeat the process on that rotation side with a smaller rotational increments. I stepped through iterations of 1/4, 1/2, and 3/4 notch positions, comparing each with the prior (I told you I was anal!).
• With this process, you should be able to find two lens positions where the focus appears about the same in both.

I finally found that I could see virtually no significant difference between the 1/4 and 1/2 notch CCW positions from the original location on my camera. It's interesting when you step through frame-by-frame to see how the H.264 compression treats the over all image, with a small area in the video appearing to go into and out of focus and back again from one frame to the next. So it becomes tough to decide which of the two clips is in better focus overall with these small lens rotational increments, but that just means you are pretty much at the sweet spot. I found three positions gave pretty much the same distant focus, with very slight differences in the close focus point. So my final lens resting spot will be based on which of three gives the better close focus. I'll finish that when a sunny days appears. I also saw that good focus in the lower left corner of my videos was almost impossible to achieve, so the cheap lens is showing it's faults there.

Good luck with your focusing efforts!

Tom,

I love your thoroughness. So, does it appear as though clockwise or counter clockwise was the right way to go to dial in your focus for Aerial video? I tried 1/4 clockwise and it was lousy. Then 1/8 turn CW and didn't like it so I reset it to original.

Also, I noted the camera guts is one tight fit in the case. It took multiple attempts to tediously place the innards in the right location so the lens was positioned correctly and case closed up tightly without forcing the screws to close it again. I did NOT have that problem with the two Basic V3 key cams.

Bill

Originally Posted by Prof100

Tom,

I love your thoroughness. So, does it appear as though clockwise or counter clockwise was the right way to go to dial in your focus for Aerial video? I tried 1/4 clockwise and it was lousy. Then 1/8 turn CW and didn't like it so I reset it to original.

Also, I noted the camera guts is one tight fit in the case. It took multiple attempts to tediously place the innards in the right location so the lens was positioned correctly and case closed up tightly without forcing the screws to close it again. I did NOT have that problem with the two Basic V3 key cams.

Bill

Thanks, Bill. When I started the focusing efforts via just the webcam visual, I marked my lens position on one of the knurled barrel ridges, not really paying attention or caring where I marked the starting location. Through numerous efforts to turn the lens with tweezers, trying to figure out if each movement improved things or made it worse, and having the tweezer grips slip off over and over, my original mark on the lens barrel got virtually erased! There is one tiny spec of paint left, but I'm not sure if it is from the right or left side of my original line. When I went to the AviDemux video method I painted a new line as noted in the valley notch so it can be rubbed off. I think my final position may be a 1/4 notch CW from where my original line was as when received, but I can't be sure. It's about 1/4 turn CCW from my second marker line if that makes any sense.

In any event, I think there's no right or wrong direction to try to focus better. But if the normal position were always slightly towards the near field as the old 808 cameras seemed to be, then clockwise rotation (moving the lens closer to the CMOS sensor) would generally move the focal point further out as we've discussed before. Your smallest test rotation of 1/8 turn was equal to 45 deg. My "one notch" initial turn was only 30 deg., with each 1/4 notch fine tuning increment being only 7-8 deg. So you may have over shot the decent focus range on both far and near sides with your larger rotations, making both look much worse. If you try it again, use the "one notch" or smaller increments.

I haven't put my case back on yet, waiting for that sunny day final test, but I think the larger capacity battery is the main reason for the new tighter fit.

Originally Posted by Tom Frank

Thanks, Bill. When I started the focusing efforts via just the webcam visual, I marked my lens position on one of the knurled barrel ridges, not really paying attention or caring where I marked the starting location. Through numerous efforts to turn the lens with tweezers, trying to figure out if each movement improved things or made it worse, and having the tweezer grips slip off over and over, my original mark on the lens barrel got virtually erased! There is one tiny spec of paint left, but I'm not sure if it is from the right or left side of my original line. When I went to the AviDemux video method I painted a new line as noted in the valley notch so it can be rubbed off. I think my final position may be a 1/4 notch CW from where my original line was as when received, but I can't be sure. It's about 1/4 turn CCW from my second marker line if that makes any sense.

In any event, I think there's no right or wrong direction to try to focus better. But if the normal position were always slightly towards the near field as the old 808 cameras seemed to be, then clockwise rotation (moving the lens closer to the CMOS sensor) would generally move the focal point further out as we've discussed before. Your smallest test rotation of 1/8 turn was equal to 45 deg. My "one notch" initial turn was only 30 deg., with each 1/4 notch fine tuning increment being only 7-8 deg. So you may have over shot the decent focus range on both far and near sides with your larger rotations, making both look much worse. If you try it again, use the "one notch" or smaller increments.

I haven't put my case back on yet, waiting for that sunny day final test, but I think the larger capacity battery is the main reason for the new tighter fit.

Tom,

I understand now. I probably did overshoot the sweet spot.
I was lucky enough to not lose my paint mark. The pudgy lipo just barely fits in the case.

I ready to order the emergency battery power accessory for $3. I bought 16 Eneloop 2000 mah AA's (the zenith of NiMh batteries) to power my DX6i and DX7 Transmitters. That means I have four extra batteries to use elsewhere. One of those would give me a lot of video time with HD camera.

Bill

Originally Posted by Tom Frank

Yes, I agree a parameter file is the better soultion. This was apparent with the old 808 #3 firmware update process to erase the date stamp, which implied there could be a parameter file involved, although the one used in the firmware update process was blank and apparently not available to the end user. I'm not knowledgeable enough in that area to know any specifics. It may require a lot more work to implement than just producing the differential firmware updates. But it has been done, I think, for one of the old 808 versions. I'll suggest this as a better solution to My HD Key Cam Guy, who seems eager to have the flexibility and features we want to see.

Different firmware versions is a quick-and-dirty solution. A parameter file, or even a USB configuration program, is the solution any serious programmer should have used. Since the original designers didn't think about using a parameter file, or maybe were too lazy to implement one, they now have more work to do. It is no great deal for a programmer to implement a parameter file as long as he/she understands the software code - recording the data is MUCH more complicated.

The parameters (just a few bytes) must be saved to the non volatile memory which is the SPI chip (the one with the 8 legs) in this case. If the processor has some non volatile memory, this could also be used. The parameters MUST be saved to non volatile memory - please make sure that your contact guy understands this. Failing to do so will result in loss of parameters if the battery is disconnected.

The #3 SPI memory is split into two parts, the boot code and the firmware. I assume that the #11 has the same mapping.

I don't know if space has been reserved for parameters in the non volatile memory. If no space as been reserved for parameters, this is a VERY BAD design, and making a universal firmware could become a bit more tricky....

Could one of you guys post a close up pic of camera circuit board...I noticed a difference between mine and the pics on chucks site...I was havin probs with mine and think I found out why...

I wanted to see whats inside the auto plug so I opened it up and slid out circuit board...I see a date but I doubt it could be used for the model of camera unless you knew they packaged them together from factory...

I happen to have my camera out of its case right now, so will post some pics of both sides of the circuit board shortly.

While your car charge is open (if it still is), can you tell us how many wires in the cable going to the camera are soldered to the circuit board (looks like just two (+ and ground) from you pic. How are the remaining wires in that cord terminated?

Originally Posted by finster

Could one of you guys post a close up pic of camera circuit board...I noticed a difference between mine and the pics on chucks site...I was havin probs with mine and think I found out why...

I wanted to see whats inside the auto plug so I opened it up and slid out circuit board...I see a date but I doubt it could be used for the model of camera unless you knew they packaged them together from factory...

Originally Posted by Isoprop

Different firmware versions is a quick-and-dirty solution. A parameter file, or even a USB configuration program, is the solution any serious programmer should have used. Since the original designers didn't think about using a parameter file, or maybe were too lazy to implement one, they now have more work to do. It is no great deal for a programmer to implement a parameter file as long as he/she understands the software code - recording the data is MUCH more complicated.

The parameters (just a few bytes) must be saved to the non volatile memory which is the SPI chip (the one with the 8 legs) in this case. If the processor has some non volatile memory, this could also be used. The parameters MUST be saved to non volatile memory - please make sure that your contact guy understands this. Failing to do so will result in loss of parameters if the battery is disconnected.

The #3 SPI memory is split into two parts, the boot code and the firmware. I assume that the #11 has the same mapping.

I don't know if space has been reserved for parameters in the non volatile memory. If no space as been reserved for parameters, this is a VERY BAD design, and making a universal firmware could become a bit more tricky....

My contact claims they developed the version of the older 808 key chain camera which implemented a parameter file. So I think they know what they are doing. But apparently there are differences with the new HD key cam making it less feasible to do at this point. They have told me a version of the HD cam that can record continuously (4 hours) is in development... don't know any more about that... may have fixed internal memory, who knows.

finster, can you say what's written on the 8-pin chip in the charger board? I can't make out anything from the picture.

Well, in the picture I see a coil, and that suggests this is a switched supply to efficiently generate 5VDC from the car battery, with switching done by that chip. But I don't see anything suggesting that any USB intelligence is there. So if this charger is special in some way, it would have to be in the way the data pins are connected.

Has anyone succeeded in charging an HD using a dumb wall wort USB charger?

PS3 controllers can be charged with dumb chargers, but the controller gives no feedback that charging is taking place. When connected to the PS3 USB it blinks, on a dumb charger it does not. I have heard that by raising the voltage on pin 2 it does communicate with the led, but I haven't tried it for fear of bricking! (they are more expensive than HD808's)

Originally Posted by wayback

...But I don't see anything suggesting that any USB intelligence is there. So if this charger is special in some way, it would have to be in the way the data pins are connected.

Has anyone succeeded in charging an HD using a dumb wall wort USB charger?

This was just recently mentioned about a dozen posts back, and why I asked about the other wires just a couple posts back!

It has already been mentioned by a user in a prior post that a generic wall wart USB charger will NOT charge this camera. I just tried it, and the red LED does not light which is the normal charging indicator. Does that mean that it really is NOT charging, or does it just mean the dedicated charger is not connected, making you THINK it is not charging? I haven't done any tests to confirm either way, but I did plug in my smart wall wart USB charger that came with my JAZZ HDV178 camera, and it DOES light the LED, which goes out after a while indicating it did put in a charge, I guess. Some tests with a dead battery would be needed to tell for sure what's going on.

[QUOTE=finster;17089753]Could one of you guys post a close up pic of camera circuit board...I noticed a difference between mine and the pics on chucks site...I was havin probs with mine and think I found out why.../QUOTE]finster,

Here's the pics of the HD Key Cam circuit board. What is the difference with yours that you mentioned? Were you able to determine how the USB data pin wires are terminated?

Originally Posted by Tom Frank

It has already been mentioned by a user in a prior post that a generic wall wart USB charger will NOT charge this camera. I just tried it, and the red LED does not light which is the normal charging indicator. Does that mean that it really is NOT charging, or does it just mean the dedicated charger is not connected, making you THINK it is not charging? I haven't done any tests to confirm either way, but I did plug in my smart wall wart USB charger that came with my JAZZ HDV178 camera, and it DOES light the LED, which goes out after a while indicating it did put in a charge, I guess. Some tests with a dead battery would be needed to tell for sure what's going on.

I have only charged mine with a wall wart charger. The one I use is a charger for a Blackberry which accepts the USB-A plug on the wall side and a USB mini-B on the device side. When I connect the two, the red LED comes on, and then goes out when it's done. After it goes out, I consistently get about 35 minutes of recording time. The Blackberry charger has a rated output of 5VDC at 500mA while many generic wall warts are only outputting 100mA to 300mA.

I have not attempted to charge this with a PC yet but I did notice the red LED coming on momentarily when I used it as a webcam. And I know my non-powerd USB port on my PC has successfully charged other devices like smartphones which have 1200mAH batteries, so they are likely rated at 500mA output also.

clx1

Originally Posted by clx1

I have only charged mine with a wall wart charger. The one I use is a charger for a Blackberry which accepts the USB-A plug on the wall side and a USB mini-B on the device side. When I connect the two, the red LED comes on, and then goes out when it's done. After it goes out, I consistently get about 35 minutes of recording time. The Blackberry charger has a rated output of 5VDC at 500mA while many generic wall warts are only outputting 100mA to 300mA.

I have not attempted to charge this with a PC yet but I did notice the red LED coming on momentarily when I used it as a webcam. And I know my non-powerd USB port on my PC has successfully charged other devices like smartphones which have 1200mAH batteries, so they are likely rated at 500mA output also.

clx1

Well.... the plot thickens!

I also used a Blackberry wall wart charger when I tested mine. I just tried it again, and saw the red LED flash for a fraction of a second when I plugged it in... might have missed that the first time. I then tried my JAZZ "smart" wall wart charger, and the PC USB port, with same results. The red LED flashes like this then goes out when the battery is fully charged. The first time I tried this with the jazz charger the red LED stayed on for several minutes before going out, so I assume it was topping off the charge during this period. And it could be the wall warts have some kind of lithium charging safety circuitry in them that terminates the charge at slightly points, making the LED act differently, too.

If I had time and desire, I'd just measure the voltage on the battery after different charge cycles with different chargers and see what the battery is doing directly from its output wires.