View Full Version : chromatic abrasion


Cameron Frittz
June 14th, 2009, 04:29 PM
for some reason with my HD-200ub im noticing alot of Chromatic abrasion with the stock lens. Anyone know what the cause could be?

Bill Ravens
June 14th, 2009, 04:43 PM
I think you mean chromatic aberration?
The 16x fujinon lens normally displays CA it hi zoom ratios and hi-contrast scenes.
Try stopping the lens to f/4, back off the zoom a bit and don't overexpose the scene.

Cameron Frittz
June 14th, 2009, 04:46 PM
Ah good to know. The scene that it displayed in it was fully open 1.4 zoomed in all the way. So that would make sense!

Bill Ravens
June 14th, 2009, 06:35 PM
Also, take great care to focus properly. Chromatic aberration increases as you move out of the plane of focus.

Bill Edmunds
June 22nd, 2009, 12:29 PM
Are there any images someone could post here that illustrate the CA issues? Are there internal settings that can be made to reduce it (gamma settings, etc)? Roughly how bad are the issues at f1.4, moderate zoom?

Steve Phillipps
June 22nd, 2009, 12:38 PM
Are there any images someone could post here that illustrate the CA issues? Are there internal settings that can be made to reduce it (gamma settings, etc)? Roughly how bad are the issues at f1.4, moderate zoom?

It's a fact of life with virtually all HD lenses, even the Canon HJ40 has it bad!
Not much in terms of camera settings you can do, but you can get rid of it to a large extent in post - not my area but I believe you'll just pick an area of CA and choose that as a colour and then de-saturate that colour, works pretty well.
Steve

Rick Bolton
June 22nd, 2009, 02:16 PM
Why would CA be more prevalent in HD video lenses as compared to lenses utilized on DSLRs? The sensor resolution on DSLRs is much higher than the 1920 x 1080 HD we are dealing with. Optics are optics - are they not? Or am I simply not understanding something basic?

Bill Ravens
June 22nd, 2009, 02:25 PM
I can take a stab at an answer, here. CA is the result of the optical error that comes from the difficulty in focusing light of varying wavelengths at the same focal plane. Modern HD(v) cameras have photo-receptors arranged on 3 different planes, for each channel R,G, and B. So, ideally, this would call for a lens that can focus the R, G and B components of the image at 3 different planes...a difficult, if not impossible, task.

Rick Bolton
June 22nd, 2009, 04:05 PM
Bill - I take it that the design you outlined is different from whatever is used on dslrs?

Bill Edmunds
June 22nd, 2009, 04:17 PM
I can take a stab at an answer, here. CA is the result of the optical error that comes from the difficulty in focusing light of varying wavelengths at the same focal plane. Modern HD(v) cameras have photo-receptors arranged on 3 different planes, for each channel R,G, and B. So, ideally, this would call for a lens that can focus the R, G and B components of the image at 3 different planes...a difficult, if not impossible, task.
Why don't I see this type of CA in the 'prosumer' fixed lenses such as the XH-A1 or something?

Rick Bolton
June 22nd, 2009, 10:27 PM
I'm not up on how dslr cameras get the R G B colors filtered to the sensor - but my Canon L glass delivers great images w/o CA that I can see - even at magnification. ED glass or fluorite elements are used to eliminate most all CA - very common in high end astronomical telescopes (TeleVue, Takahashi, etc.).

My point is that good optical design exists for solving this.

Any updates from anyone on the Canon 14x for the 700???

Steve Phillipps
June 23rd, 2009, 12:54 AM
Why would CA be more prevalent in HD video lenses as compared to lenses utilized on DSLRs? The sensor resolution on DSLRs is much higher than the 1920 x 1080 HD we are dealing with. Optics are optics - are they not? Or am I simply not understanding something basic?

It is there in DSLR lenses, more so in some lenses than others. It's certainly there in prosumer fixed lenses too.
Steve

Bill Ravens
June 23rd, 2009, 06:26 AM
I beleive most DSLR sensors use a single matrix of photoreceptors. Unlike videocam 3-chip sensors they are all located on the same focal plane.

IMHO, most of the CA is the result of inadequate optics design. I say this because my stock 16x and 17x Fujinons, as used on the JVC HD110 show very high CA, especially at extreme zoom. I am also using several different Nikon 35mm zoom lenses, couple with an MTF adapter for telephoto shots on my HD110. CA, is virtually un-noticeable with the 35 mm lenses. Why this is the observable case, I can only assume that it is the higher quality of the 35mm glass. Another possibility: high zoom ratio lenses are harder to optimize than low zoom ratio lenses. Video zooms like the 16 and 17x cover much more of a zoom range than the 35mm equivalents, which are more like 8x. As I said earlier, the appearance of CA is highly related to out of focus areas of the image. 35mm lenses have a much larger circle of confusion, so, the apparent CA is not as noticeable. The shallow DOF of 1/3 inch lenses tends to amplify the occurence of CA.

edit: On second thought, this is a bogus ratonale because my 35mm lenses maintain their CA-free performance, even when used on a 1/3 inch sensor.

Steve Phillipps
June 23rd, 2009, 06:57 AM
As I said earlier, the appearance of CA is highly related to out of focus areas of the image. 35mm lenses have a much larger circle of confusion, so, the apparent CA is not as noticeable. The shallow DOF of 1/3 inch lenses tends to amplify the occurence of CA.
.

There is a contrary line of thought that the wider the aperture the worse the CA looks due to the fact that the fringed borders cover a wider area due to being out of focus. This makes sense.
I think you're right in your evaluation fo why it's there - video lenses stretch optical design to - and it seems beyond - the limits of what's possible, while your DSLR lenses are much simpler by comparison. Also the HD lenses need to be designed to resolve much finer detail due to the smaller pixels on the sensors of 2/3" and espcially 1/3" cameras.
I'm no scientist, just my thoughts and observations.
Steve

Bill Ravens
June 23rd, 2009, 08:22 AM
35mm lenses have a much larger circle of confusion, so, the apparent CA is not as noticeable. The shallow DOF of 1/3 inch lenses tends to amplify the occurence of CA.


Thanx steve....this whole statement is wrong. I got it backwards. I.E. 35mm lenses have a shallower DOF(smaller circle of confusion) than 1/3 inch lenses, but, it's all irrelevant, anyway, when using a 35mm lens on a 1/3 inch format.

Rick Bolton
June 23rd, 2009, 01:36 PM
OK - I'm going out on a limb here and may well display my ignorance on the technical realities - flame protecting suit is on.

Canon (dslr) and some others produce stunning high resolution still images with their optics and capture block. They sport a FULL 35mm sensor with FULL CONTROL capturing resolution way way beyond 1920.

Enter HD video with the need to capture 24 frames per second for extended clips and go to larger ccd sensors or faster cmos to avoid jello vision. The larger zoom ratios may be causing the problem (my dslr are like 3x at most) - perhaps we need less zoom and better CA performance.

Why can not SONY or Panasonic get it all together in a package? Might that prosumer ($7,000 give or take) camera eat into their much higher end cameras and erode their corporate profit margin? I think so. If the prosumer end gets much better there may be far fewer willing to pay the ultra high premiums for the pro line. Thus, we appear to get "compromised " releases. I fully believe the capability more than exists to manufacture global shutter cmos or to cool larger ccd sensors or to produce CA free zoom lenses at 10x or so.

Alex Humphrey
June 23rd, 2009, 10:01 PM
well I always said for the last 20 years to pick the lens with the less zoom range, you tend to get a better image (given same brand and price range) with a lens that doesn't try to do as much.....

Lenses are a combination of compromises. Fixed focal length lenses have some compromises and zooms have more. The more zoom range, the more compromises. Though a $8,000 lens usually looks killer by anyone's yard stick regardless of range... At that price manufacturers rarely make lemons.... rarely....... (why do I think of Aliens... "they mostly come at night....... mostly...")

CA exists EVEN on a fixed focal length $3,000 35mm lens.. you just don't see it easily.

I would prefer to have a shorter zoom range, or even just a digital readout in the viewfinder to remind me of my range so I know where not to pass in my range if I get distracted.

Rick Bolton
June 24th, 2009, 11:41 AM
Alex - '... you just don't see it easily ..." - sooooooo important.

I've been using a lil Canon HF 100 to educate myself into HD video. To some extent, the pixel peeping, tech specs, and internet video samples have both educated me and clouded my decision making (or lack there of) relative to the next HD video purchase.

When I took the SDHC card from the HF 100 and played it through our Panasonic BluRay on our Panasonic Viera - WOW - stunning clarity / color / detail. Several of the locked down clips would stack up with the best I've seen from Discovery HD or Animal Planet HD. That is NOT a comment on any skill I may have but rather what this equipment is capable of - even on auto function.

My plan is to wait a bit longer to see what Canon comes up with next - and to see the reviews of the Canon 14x on the JVC 700.

At days end it is about the image / composition / flow / planning / .......