Pentax Q-S1 using C-Mount Lens Adapter and Lenses

[VidThreeNorth]

I bought a Pentax Q-S1 recently specifically to use with C-Mount lenses. Right now, I do not have any actual "Q" lenses -- just C-Mount. This is not a "primary work camera" so I am only using this camera occasionally when I have an opportunity. I have gone out of my way to do some testing just to make sure everything was working, and to begin familiarizing myself with the equipment, but I am not in a hurry to master this setup.


First Project

These are early equipment tests. I am testing specific setting which are not necessarily optimal for the situation. At this stage I am, just finding how various settings interact.

The main focus of this project is to test the performance of Ricoh FL-CC2514A-2M 25mm F/1.4 (2MP rated) C-Mount Lens at F4.0.

It was also a test of the "Bright" setting in the camera, along with the "Highlight Compensation ON" setting.

This set of pictures was also my first set done with this camera and a monopod. Everything else so far has been handheld, though some cases with my body "braced" against something stable.

In this project the lens was locked at F/4.0 for all pictuers. Best resolution for this lens is expected to be roughly in the range of F4-F8. So this should give some of the sharpest possible pictures, but at the shorter end of depth of field, though with expected minimal motion blurring, and best performance of the sensor. I should mention that the only serious problem I have with this lens is that the aperture ring moves a bit too freely. Because the lens is so small, the aperture ring can shift when I am focusing if my fingers touch it -- which happens often. I have used an adhesive tape to "lock" the aperture, but that has not worked well. I might start using the supplied "thumb screw" to lock it, but that screw was not meant for handheld use and is cumbersome.

The "Bright" finish used in this set of pictures is the default setting for this camera. The metering is center-weighted.

For Pic01 the dynamic range is "standard". In all three images, the lighting has a problem with highlights. The Q-S1 warned me about the highlights, flashing them red. The Q-S1 has an "HDR finish" but I am not trying that in this project.

In Pic01 the situation was not too bad, so there was a chance that the image would be acceptable. In fact, I think a lot of people would be happy with this as-is. Personally, I think I might have preferred a bit less contrast, and 1/3 stop darker to bring down the yellow of the hanging bars. But going darker without decreasing contrast would would have blacked out the foreground shadows and the shadows on the trees. ISO was 100 and exposure time was 1/500th sec., implying plenty of room to make changes.

In Pic02, there was one particularly bad highlight, on the rock near the lower-middle of the picture. Again, the camera warned me about this one. I turned on the "Highlight compensation" in the finish, but still got a warning. I could have done more, but I decided to take the picture and see how it turned out. The result is still too light (and clipped). Here again, lowering the exposure would have pushed the shadow areas, particularly in the right side, down into black, so the next step would be to either use "Shadow compensation" or switch to "Natural finish" each of which flattens the contrast, and then perhaps to decrease the exposure. Even with that done, I think it might have been better to recomposed the picture putting that rock in the foreground. I do not know how that would have worked out. If I am ever back that way again, maybe I can try it.

Pic03 was a similar lighting issues to Pic02. Again, the "Q-S1" warned me about the highlights. Again, I limit my adjustment to turning ON the "Highlight compensation" but without any further setting changes. Again, I think a better picture was possible by most likely by adding "Shadow compensation" and dropping the exposure, probably about -2/3 stop.

As for the lens, at F4, despite its "2MP" rating, it did a very nice job. Ricoh claims that their "5MP" lenses have more even lighting (less drop off from center to corner), and of course, higher general resolution. The VGA lenses are noticeably too soft, and are only good enough if you want such a soft appearance. The Q-S1 estimates the "35mm camera lens" equivalent of the 25mm focal length to be 115mm which is a short telephoto. Having lenses like this around $100 - $150 US is part of the advantage of using C-Mount. In theory, the other advantage is that it should save some battery life. Without the need to power mechanisms for auto-focus and auto aperture, a battery change should last noticeably longer. I would like to prove that someday, but it could be difficult.

Future Plans:

Aside from getting some general photography done with this camera, I would also like to do an F8 day or an F2 day to see the depth of field and lens performance differences.

Picture Notes: The forum does not allow the full size 4000 x 3000 pictures, so I am uploading 1200 x 900.
[2017-07-19 21:34 EDT, I removed the 1200 x 900 versions that I uploaded when I started this thread, leaving the 1640 x 1230 versions in their place.]

About the 1640 x 1230 size:

As I wrote in post #3:

"... Real "2 Meg" is actually 2,097,152. So, in order to better represent the "2MP" lenses, for today, I will try to use a 1640 x 1230 image size ( = 2,017,200 pixels). If you download and look at the images in these sizes, you can judge whether the have met their intended specifications. In fact, I feel that this Ricoh lens has done so."

I have made 1640 x 1230 versions of the pictures I originally uploaded for this post. I will leave the 1200 x 900 versions for a short while. I do not see any reason to leave them, so I will eventually delete them and rewrite this post to refer to them. If you have a good reason for leaving them available, mention it here and I will leave them, but I think they are just clutter with the new 1640 versions available.
[added 2016-06-30]

Wall:
Pic04 (set includes "b" bottom left and "c" center)

I am including these images in in post #1 because I used F4.0 for them, and so that is where they fit best. EXIF data says 1/250 sec., ISO-100. Highlight and shadow compensation is Auto (probably not used for this set of images) and finish is "Natural".

I thought that I should add an image set of the wall that is shown in Post #3 using the Ricoh 25mm lens. This set of images cannot be conclusively compared with the other set. I was using not able to use a tripod for either set due to problems caused by the C-Mount adapter (which over-hangs the bottom of the camera body). I was only a bit more successful using my monopod. But a monopod is not a complete stabilization. It moves around. I hand focused the images and shot with some slight motion present. I made multiple attempts and I am working with the largest file which probably is the file with the most detail. I have another C-Mount adapter on order but it has not arrived yet. However, when I get it, I doubt if I will bother re-doing these tests. It really is not as scientifically definitive as I would like. But as a crude test it can confirm if there are obvious problems.

In this case, there is not a lot of surprises. In the 1640 reduction of the whole image, linearity is pretty good (the horizontal line of the top and bottom rows of bricks are fairly straight). The wall is a bit dirty, and lighting might not be perfectly even, but you might notice a slight fall-off of brightness from the center to the corners. But it is so slight that you might convince yourself it is your imagination.

As I have noted often, I think that the sharpness of the lens reasonably fulfills its declared specifications. If you look at the 1:1 scale "Detail" images ("B" bottom left corner and "c" center), you can easily confirm that the corners are not as sharp as the center. But if you look at the 1640 reduced image though, the sharpness fall-off is not bad. If you draw a circle in the middle with a radius of about 80% of the distance from the center, and cut the corners off at that distance, that would pretty much cut off the noticeably "soft" areas.

As for how far beyond 2MP it can be used. All I can say is that it depends. It is not just a matter of being "fussy". One can have specific objectives where a softer lens might be adequate or even preferable. For use on an action camera capable of "4K video" I think it is borderline. I would prefer to try a 5MP rated lens if I had one. But it is not so bad that I would go out and buy a 5MP lens right now to complete a video -- not unless someone else was paying for it. . . .
[added 2017-07-01]

 

[VidThreeNorth]

This second set of pics was taken as an "F8" project. All the pictures were taken at F8-F9.5. The aperture drifted a bit because the cloth tape I used to hold it stretched. From now on, I plan to use a combination of tape and the supplied locking screw to hold the aperture ring in place.

NOTE: At around F8 and smaller refraction is probably occurring and sharpness might actually decrease though depth of field is increasing. I would have to do a more formal test to tell for certain, and frankly I do not expect to have the time to do so in the near future.
[2017-06-29 added this para.]

This session backfired. I went to a park mid-week at around 15:00 expecting it to be fairly empty, but because the weather has been so rainy this year, I think a lot of people went out of their way to get out in the sun. There were a lot of people around. Trying to avoid people in the pictures forced me to work closer to the birds than I wanted, and their natural desire to avoid me meant that I had to rush.

The settings were "Natural" finish, with both "Highlight" and "Shadow" compensations on "Auto". I have tried these dynamic range compensations in forced "On" and "Off" before so I think I have a good idea of their impact, which is mild -- probably about 1 stop for each. For this session I followed the "overexposure warning" (flashing the over exposed area in red) and compensated downward. F8 makes focusing more of a challenge. I think I am finding I prefer to work at around F4 - F5.6.

Pic01:
As best as I can recall, I focused on the duck in the foreground and then composed with the goslings in the background. At F8, the depth of field covered the background about what I expected. Overexposure compensation for Pic01 was EV="-1.0". No adjustments were made in post. The "detail" crop I am uploading should be 1600 x 1200 (same as the reduced version of the whole image) is original magnification and should show accurately the lens' characteristics. Exposure was 1/160 sec, ISO-100.

Pic02:
For this picture, I only adjusted part way for the overexposure warning. I dropped down EV=-2.0, but there was still a bit of "flashing red" on the tail end of one of the geese. I was hesitant of dropping further due to the already dark shadows. I think that much of the picture worked well enough. I am posting the image reduced out of camera but without further adjustments. This picture actually cleans up nicely in Corel Paintshop Pro X9 "SmartFix", and likewise in some other finishing software. Exposure was 1/250 sec at ISO-100.

The question here is whether there is much point in spending the money for the better "5MP" lenses? It is very borderline. Yes, I think I can see room for more sharpness, but then again, most of what I record is video, with movement. These "still" picture have live animals which were moving around. A beginner assumes that a measurably sharper lens always give a sharper picture, but motion blurring is more of a factor than people realize. With so much moving around, would I have gotten more detail? Maybe the grass and such might be a bit sharper, but the living, moving birds or blowing tree leaves? I doubt it. Still, within the range of "2MP" and "5MP" lenses and 12MP sensors and images, yes, I think a good case could be made for paying the extra money for the better quality. Maybe someday I might do that, but not right now.

Also, I think that a "new Q" could sell well. I would aim it at the video field with 4K support. I am not sure about what sensor I would recommend. There is a constant push for more pixels, but the quality of image at 12MP is quite nice. I would certainly not go higher than 16MP, but I would go further and say that I would probably recommend around 14MP, and aim for smooth image quality and better low light than a 16MP sensor would provide.

 

[VidThreeNorth]

This time: two lenses

I am finishing up my tests of the Ricoh FL-CC2514A-2M 25mm 1.4 with a test at F2.0 and making a short comment about a Computar M0814-MP2 8mm F1.4 (also 2MP rated) lens.


About the 1600 x 1200 size:

I was not really paying attention to this before, but it occurred to me that 1600 x 1200 was about 2MP. I checked it and found that it is actually 1,920,000 pixels. This is actually the size range the lens is intended to produce. Real "2 Meg" is actually 2,097,152. So, in order to better represent the "2MP" lenses, for today, I will try to use a 1640 x 1230 image size ( = 2,017,200 pixels). If you download and look at the images in these sizes, you can judge whether the have met their intended specifications. In fact, I feel that this Ricoh lens has done so.


Ricoh 25mm, F1.4 at F2.0

Big F-stops are often used for depth of field control. My first 35mm SLR film camera only supported "stopped-down metering". That is to say, that in order to meter, I pressed a button and the aperture closed down to the selected F-stop and then the meter turned on. I did not necessarily know exactly what my F-stop was, but I was aware if it was fairly open or closed, and I had an estimate of my depth of field plainly visible. I think I was lucky. Cameras like that are probably the best to learn on. My later cameras supported "open-aperture metering", but I had already learned depth of field control. I think people who learn photography on these "better" cameras probably have more difficulty learning depth of field control.

Today, when I record video, I rarely deliberately use depth of field control. I usually try to get everything in focus. Zoom cropping is more often the technique I use to direct attention. In my videos, a blurred background more often comes from panning to track motion. If I am working in 3D, then I emphatically want everything in focus because it is easier on the eyes that way. So no, I do not generally use depth of field control.

I like having "big F-stops" available to support shorter exposure times which helps reduce motion blurring, and lower ISOs, which reduce image noise. In general, you will find me using apertures bigger than F3.5 in evenings or dark environments -- and frankly, not liking it much.


The Pictures:

All the Pentax sample pictures were originally 4000 x 3000 pixels. The picture of the Pentax camera with the Computar 8mm lens was taken on a Sony a5000 and cropped, labelled, and resized.


Ricoh 25mm, F1.4 at F2.0 set:

Pic01 (0219)
- 1/250th sec, ISO-100, Exposure -2 step, w/Monopod
- original file size 1,726,742 bytes (small), Settings were "Natural" w/highlight and shadow compensation Automatic, High ISO noise reduction "Auto".

Re-processed in Corel Paintshop Pro X9
SmartFix:
Brightness
Overall 26
Highlights 10
Focus 45
Black 2
White 22


Was this picture a failure? I was looking for a location for a sundown/sunset video and decided to use the opportunity to continue my exploration of the Q-S1, 25mm F2.0 setup. This scene was a "throw-away". I gave it some thought, and but it was not really what I was looking for -- either for the video or for this set of stills. It was later, when I was going through the images I noticed some interesting aspects -- in particular the division of the smooth textures of the sky above the rougher textures below of the trees, and the urban hint of the telephone lines and the buildings on the right.

This is the first "doctored" picture I have uploaded. I was hoping for a bit more detail in the foreground. The EV = "-2.0" was used to retain the sky colour, but too much of the foreground detail went away. I have a DNG file for this image and even that does not help. I think the only real solution was to use the built-in HDR capabilities of the camera or make an exposure bracketed set of images and "hand craft" it. From what I understand, if I use the built-in HDR settings no DNG files are saved. The camera processes a single JPG file and that is all you get. I have no experience hand making an HDR picture, but I think I would probably trust my luck to that approach. So a set of bracketed exposures would be my approach.

This image was re-processed from the JPG in Corel PaintShop Pro X9. In effect Paintshop adjusted it to about 1 F-stop brighter, which is what I should have used in the first place. I brought it down 2 F-stops on the advice of the Pentax' over-exposure warning.

[2017-0629
I have uploaded a new version of Pic01. After spending more time on it in PaintShop Pro X9 I came up with a version that is closer to what I wanted. Working from the JPEG, this time I used "Histogram Adjustment - Midtone Compression 20" and then Smartfix:
Brightness Overall 7 (recommended was actually 28)
Highlights 10
Focus 60
Black 4
White 26

This is the closest I can come to what I really wanted from this image. As I noted before, I should have paid more attention to it when I took this picture.]

I focused on the silhouetted trees that are about 20% from the left side of the frame with the hope that the lower trees in the foreground would have enough exposure and detail to provide an interesting offset to the smoother sky. At F2.0, the depth of field is shallow. Some trees further away are soft, as are the buildings and telephone wires in the far right.

The leaves in the far left tree are only a little out of focus. When I enlarged the camera file, I found that they are also motion blurred, possibly due to a local breeze.

I would be happier if the Q-S1 had a higher JPEG quality setting. The current top setting does not seem to allow a 12MP image to create a file larger than 7MB (the "brick wall" pictures) and most "real" picture are below 5MB. That is too small. I have already had pictures with noticeable artifacting. They should add a setting that peaks higher. Also, I would prefer having a PNG file option. The semi-pro and pro crowd loves DNG, but I would generally rather use the built-in processing, but without the "lossy" JPEG storage.


Pic02 (0223)
1/2000 sec., ISO-100, no exposure bias, handheld, "Natural" finish original file size 2,817,397 bytes

This is more typical of what some people would do with a short telephoto and a large F-stop. Actually, I would have preferred an F3.5-4.0 composition, but today was my F2.0 day, so...

I have grown geraniums, but I am not really a fan of them. I do like the magenta varieties. The magentas have more complex colours in the petals compared to the reds. The Q-S1 has done a good job capturing the subtle tones. I have included a full magnification "detail" crop of the main flower and as I have noted before, in the center, there really is not much room for improvement for a 5MP lens. I would expect that corners might turn out better, but in the larger F-stops, depth control style pictures usually leave the corners fuzzy anyway.


The Computar M0814-MP2, 8mm F1.4 lens, rated for 2MP
- all pics @ F4.0

Computar lenses seem to be popular lenses for security cameras and industrial "factory automation" applications, so I was interested in seeing what kind of quality can be expected in their products. Lately, it seems that these lenses have come up in price, presumably due to the interest among the action camera conversion users. The lens I bought was not one that was important for my needs. I am well covered for wide-angle photography, even just using my cell phones, and lately this lens costs around $140 US. It does expand the Q-S1's abilities, but my main reasons to buy it were that I have been thinking about using this on an action camera C-mount for video, and it was a chance to see what Computar products were like.

A Broader Perspective:

Pentax' "Q" lens array lists their "01 Standard Prime" lens as an 8.5mm F1.9, which comes as a surprise for me since its FOV is only a bit narrower than this 8.0mm, which I consider a "mild wide angle". The Pentax lens includes open aperture metering with full automatic setting and an internal "leaf" shutter function, all for about $200 US. The only reason to buy this C-Mount lens at all is to be able to share it with a "4K" capable action video camera.


Pic03 (IMGP0232)
- set includes "b" Bottom Left and "c" Center crops.
- 1/200th sec., ISO-100
- original file size 6,129,193 bytes (one of biggest I have seen from this camera)

Center sharpness is good but falls noticeably off off-axis. In some pictures that I took on a Sony a5000 there were indications of misalignment. Part of the picture appeared to have barrel distortion and part had pincushion distortion. This picture does not show that problem clearly. Chromatic aberation appears beyond ~25% off-axis, which includes most of the picture, but I do not find it severe. But there does seem to be a problem with this lens. If you look at the upper right corner of "Pic03" it is noticeably softer than the other corners. When I rotated the camera 90 degrees in "Pic04", the upper left corner is noticeably softer than the other corners -- this is the same part of the lens. It will appear in "Pic05". This seems to confirm that something in the lens (probably a lens group) is misaligned.

About the "tilt": The "vertical" groove in the wall was not truly vertical, so I had a choice of either making it look vertical by tilting the camera or making the bricks look level (which they are) and having the groove tilted. I actually did both, but this this file was largest, which indicates it probably had a bit more detail. So this picture is not necessarily the "prettiest" but it is the best functionally.

Pic04 (0247)
1/640th sec, ISO-100, F4.0

Pic04 was taken later but using the same setup. I did not remove the lens from the camera between the pictures. It is important to know this because C-mount is a screw mount, so the lens alignment can change from one mounting to another. While handling the camera I might have changed a setting affecting the final processing. I think I turned on "tone expansion". At least it was set that way later when I was charging my battery. The "nicer colours" are at least in part by better lighting on the second attempt, but might also have benefitted by this setting.

See my comments about Pic03 regarding misalignment.


Pic05 (0212)
- 1/250 sec, ISO-100
- exposure bias "-1/3 stop"
- original file size 5,650,689 bytes (typical for this camera)
- finish was "Portrait" which has slightly exaggerated saturation which works well in this picture.
- Highlight and shadow compensations were OFF.

Three corners are not too soft, but the upper right is bad.

This picture in its "1640" resized state hides the softness in the upper left and lower corners fairly well by the composition. The grass and leaves in the bottom left and right look properly soft as being at the limit of depth of field. The leaves in the upper left are similarly excused, though if you check around the middle of the picture you will finds trees a bit further away that are sharper. The upper right corner looks badly blurred.

This is a "perfect storm" situation. The generally softer corners of this lens combined with the apparent flaw of this particular sample, has combined with the trees being beyond the depth of field, and the type of trees (some of them appear to be conifers) which have needles instead of leave, which are never going to look sharp even at moderate distances, all combine with this result. The only good news about this problem is that there is a 2-piece C-mount adapter available for the "Q" cameras that allows re-orienting a lens with a problem like this to put the "soft-spot" at the top or bottom of the frame, where it will affect as much of the picture frame. But I do not have that particular C-mount adapter.

Where was the focus? Actually, I am not certain now, but I think I focused on the tree to the left of the foot-bridge, with the fine bark pattern, or the dead tree trunk beyond it which has nice colours. Either way, the depth of field covered the foreground plants and fairly far into the background (perhaps around 100M), but not to infinity. The composition feels limited. In fact, it looks like it could have been staged indoors with a matte background.

Pic06

The manufacturing of the Computar lens is not as cleanly done as the Pentax. Aside from the dust that is on the camera when I took it out of my camera bag, I have indicated where glue was visible around the retaining ring on the front of the lens.


Re-Think:

First, I have mentioned prices vaguely so far. In general I have found that the 2MP quality lenses seem to sell for ~$150 US. while the 5MP lenses seem to sell for $250 - $300 US. For the price of my Ricoh 25mm (2MP) and Computar 8mm (also 2MP), I could have bought one of either a 25mm or 8mm lenses in 5MP.

I can say that if you are considering a GoPro Hero4-5 or Xiaoyi 4K or 4K+ for "4K Video", and conversion to "C-mount" then you should invest in the 5MP lenses. Those action cameras use 1/2.3" sensors which are just slightly smaller than the original "Q" (1/1.7") sensor. The 2/3" sensor on my Q-S1 already shows that it could benefit from the 5MP lenses, so there is not doubt that the extra quality is going to make a big difference on the smaller sensors.

 

[Derrel]

Good write-ups on your endeavors. Thanks for posting this, and with all those details, we know plenty of effort went into this. Keep up the good work!

 

[VidThreeNorth]

Thanks for your appreciation! It was an unusual effort for me to write these up. There is probably an interest in the lenses from the Action Camera crowd, but as far as we can tell, the camera is probably history now, so there may not be as much interest there. It means in effect that everybody who is going to have bought this camera already has, and in fact, I am probably one of the last. So I will probably not be saying anything that is not well known to everyone except me.

I also expect that I will have questions that other people might be able to answer, which will probably help me.

I have cleaned up grammar and formatting (just before I posted this message) in the previous posts and in the coming weeks, I will probably post a few more pictures if I find any that illustrate some points I have not covered.

And yes, I am thinking about buying another lens. But I really should be doing other things right now, so I have no immediate plans right now.

 

[VidThreeNorth]

My schedule has been difficult this year, so this was my first change to return to the second location from the F2 image sets. I was hoping to get another try for more sunset/sundown, but the weather did not co-operate. It was only around 50% overcast, but the clear patches were high in the sky where they did me little good.

I took the opportunity to experiment with the HDR setting and came away with some pictures that I like.

Ricoh 25mm, F1.4, 2MP ("FL-CC2514A-2M")

These are more pictures at F4. All pictures were taken with a monopod. Due to the C-Mount adapter I am using, I cannot use my tripod. I have another adapter on order.

A Question of Priorities:

Learning a complex camera means dealing with conflicting settings. Some settings block other, but sometimes settings are prioritized. The HDR setting in the Q-S1 is blocked completely if your output is anything but JPEG. HDR cannot be set at all if the output is RAW or RAW+. The camera must be changed to JPEG output before the menu will allow access to the HDR setting. On the other hand my basic settings lately are: Custom Finish Portrait, with Highlight and Shadow compensation both Auto, and High ISO noise reduction on Auto, as well as the ISO and Shutter Speed.

But when HDR is active, it appears that the "Custom Finish", and "Highlight" and "Shadow" compensations are de-activated. The settings are still there, but they do not seem to do anything. The HDR seems to override them -- HDR has a higher priority. However, I was able to change the EV compensation with HDR on, and when I did, it seems to either re-adjust the exposure(s) that would have been set by the HDR function, or to take the exposures and change the result. I will need more testing to know which is going on, but that is what it looks like from these pictures.

Without a known reference (these settings are not all covered clearly in the manual), it could take me a while before I have these worked out.

Pic01 (IMGP0274)
HDR ON, 1/125 sec., ISO-500

This picture is a fairly subtle HDR. The only reason I know it, is because I took the picture myself. If someone told me they took it with a camera which simply had better dynamic range, I do not see any way that I would be able to say otherwise. Some people would prefer this picture with further adjustments, but I like it the way it is.

Pic02 (IMGP0279)
HDR ON, 1/125 sec., ISO-100, EV -1 step

This picture looks like there might be mountains in the distance, but in fact it is just clouds. I think this looks a bit more like an HDR than Pic01. It looks more blended around the silhouetted trees, though even here, if I were shown this by someone else, I do not think I would spot it.

Pic03 (DSC00067 - replaced by DSC00072)
The Ricoh 25mm F1.4 (2MP) lens at F4.0 has an interesting aperture shape. It is a 5 pointed star, but with accentuation. [Taken with a Sony a5000 and cropped. The auto-focus missed, but I did not realize it at the time.]
[2017-07-12
I have replaced the picture of the Ricoh 25mm lens aperture with a newer version that is properly focused (manually of course). Again, this picture was made with a Sony a5000 and cropped and reprocessed in Corel Paintshop Pro X9.]


Pic04 (IMGP0300)
HDR OFF, Finish Portrait, Highlight and Shadow compression Automatic, 1/125 sec, ISO-1600, EV -0.7 step.

The obvious thing to notice about this picture is the flair patterns on the headlights. Somewhere, I think I still have 4 point and 6 point star filters, which I like, but are only useful in limited situations. These flairs are a bit more natural, but they do add interest.

Aside from cropping, I did "obscure" two license plates. There was just enough image that I could read them if I made some effort.