HOYA PRO1 MC UV Review + Filter vs No Filter Tests

Filter: HOYA PRO1 Digital MC UV(0) ~$44-$64 US at time of review (Mine was $44)
THK Photo's Site – Link to Product
http://www.thkphoto.co​m/products/hoya/gf-01.html

Canon Rebel XT & Canon EF 70-200 F4 IS @ 138mm


Lighting: Approximately 250w of 5500K light (umbrellas) – All ambient light was turned off.

Testing: Four pictures were taken with and without filter to examine color rendition and contrast in a controlled environment. The “Family of Angles” was adhered to to reduce 'any' reflection or glare within the scene. I picked from the four pictures of each judging on sharpness. All shots were treated identically; one was adjusted for -1/3 EV in Bibble by .33, sharpened by a factor of 1 in accordance with Canon's recommendations (the Rebel XT is by far not as sharp as the 5D or 1D Series), cropped to 8x10, all other settings were unchecked or deselected and the exact settings were applied to the second picture. The full pictures were resized (Bicubic Sharpener) but the 100% crops were saved as JPEG's, posted to Photobucket and then posted here. In truth, the provided pictures are just a formality since the only true data is in the color tests done within Photoshop and I worked exclusively with 8-bit .tif files. They do however provide you with the scene and the range of colors. Please forgive the noise in the photos as I had to shoot at ISO 400 to get the speed up as I do not yet have a tripod ring and the camera did shake enough to blur the pictures at 1/8 (where the lens sweet spot is at 5.6 @135mm). The pictures actually did come out better with me just hand holding with IS on, but to keep it more scientific I used a tripod.

Without HOYA PRO1 Digital MC UV(0) – The Control

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

Pt. II

The first test was conducted in a controlled setting and did show some difference between with and without filter. There was a slight increase in noise and slightly less saturation of colors. The noise was negligible and unless one was pixel peeping it would go unnoticed by most people. The very slight lack of saturation was constant in all four shots – I could see the difference immediately. However, this can be compensated for in post processing and again, most people would not be able see the difference unless they were told what to look for. Detail however stayed constant – the filter did not seem to effect detail at all.

The second test is a real world test of the filter and is aimed at simply comparing contrast and color. No measurements will be made because there simply is no way to set a 'control' and thus, all conclusions would be highly subject. Again, the photos were treated equally; setting were made with one then pasted to the other so they would be identical. (No hood w/ sun behind me behind clouds, 200mm, ISO400, Cropped to 8x10, EV -1/3, Bibble +.33, WB Sunny, Sharpened +1 and all other settings at 0 or unselected).

Without Filter F5.6 1/160

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

HTTP response: 404 | MIME changed to 'image/gif' | Byte size: ZERO | PHOTOBUCKET ERROR IMAGE

I found no image quality difference when I tested my filters. hoya HMC, hoya S-HMC, Hoya Pro1 digital, hoya pro 1 CP, Hoya CP, kood P-series CP.

All filters will degrade the ultimate IQ to some degree. The question then becomes does what the filter add e.g. cutting haze for UV or reflections for a polarizer provide a greater overall benefit to the image. In some cases yes but not all. I currently have Hoyas and am looking to upgrade to B+W. Will see how much of a difference that makes.

Ya, I am surprised there are not more definitive tests and reviews comparing no filter versus the various filter and protector brands. I agree, there are times when wearing a filter is best for the image. But most of the time, simply removing the protector will get you the best possible image. It only takes 10 seconds to put it back on.

I don't really notice filter quality so much when viewing on a monitor, but when printing large photos, all filter and lens weaknesses tend to become exaggerated. And to me, the final print on the wall is ALL that matters.

OP, thank you for posting this. I am fairly new to digital and have the same lens and filter that you used. The results suprised me!

Eye opening .. I've not seen many filter tests and had always presumed the effect of quality filters to be negligable, but that is not the case here.

In the interest of complete control the outdoor test seems to have foilage in the BG which may have moved slightly (and thus affected the reflectvity). I'd like to see one with a constant source of light causing the flare.

Regardless, the effect of these filters is more than I thought. Thanks for showing this.

Just curious, for the first tests, you say that you sampled using the grid in PS, but did you use a single pixel to get the color or did you have the picker to select a bigger square of pixes, like 3x3 or 5x5? Since there is noise, I would think that the 3x3 or even better the 5x5 would give a more accurate assessment than a single point.

A truly scientific test would involve taking a randomly distributed sample of shots taken with and without the filter (known only to one person helping with the experiment) and then presenting all the photos to an observer who has the challenge of identifying the filterless vs filtered photos. Then if the statistical correlation between 'identified' vs. 'real' is high enough, one can then draw the conclusion that filters do alter images to an identifiable degree.

Filters increasing noise and affecting bokeh? I don't think so.

They might change colors slightly (if the filter has a cast, for example) or saturation (if there are reflections or flare causing loss of contrast).

The color sample size you took seems too small to me to give meaningful color values even with the grid, etc. like TTSkipper mentions.

If you check out the Protective Filter FAQ (link in sig), you will see that others have been before you.

Comments on your results:

1. Colour Tests: Yes most UV filters will introduce a small colour bias. Look at Q16: "In the case of a Hoya SHMC Pro 1 UV(0) there was a small warming towards red of about 2 steps in 210 (red-blue) or 100K in 5250K in terms of colour temperature. ", the answer includes two small swatches. Not a big issue specially if you take a WB reference.

2. Contrast: Look at Q8: Detailed tests under controlled conditions with in frame relative light source levels about the same level as the naked sun; showed no obvious effect on contrast and for most lenses minimal additional flare and ghosting, the lens being by far the dominant factor.

An exception was deeply recessed lenses where the filter could introduce some ghosting under these very harsh conditions. My concern with your method here is being outside the conditions are not well controlled.

3. Sharpness: Look at Q8 again. This has been tested using SRF/MTF test methods, the addition of the filter made no statistically discernible difference.

Note however, that measurements to this accuracy did show significant camera AF accuracy from shot to shot with or without filter.

Unfortunately this means it is very difficult to measure for any effect on sharpness simply by just comparing two shots, the AF accuracy is just not good enough for that: For that reason this is one of the most common areas that testers have systematic test design problems.

4. Bokeh: I'm not sure what you are seeing here, perhaps you could markup a shot with arrows or circles to show what areas you are referring too. I can see some small difference in what looks like flare, but in uncontrolled conditions this could be due to movements of sun-blocking vegetation or movement of the sun itself.

Now it seems that there can be a very strange double image effects in out of focus areas when single coat filters are used (Again see Q8 for discussion but at this point no examples)

I have checked for this with Hoya SHMC Pro 1 and found no indication of this although I don't have the examples up at present. I have never heard of bokeh problems being claimed for high end filters prior to this.

I'm glad this has been informative for people. I have purchased more filters to test; Hoya's and B+W (ouch that was expensive) and will be doing more tests once I have the time.

mmahoney wrote in post #4496367
Eye opening .. I've not seen many filter tests and had always presumed the effect of quality filters to be negligable, but that is not the case here.

In the interest of complete control the outdoor test seems to have foilage in the BG which may have moved slightly (and thus affected the reflectvity). I'd like to see one with a constant source of light causing the flare.

Regardless, the effect of these filters is more than I thought. Thanks for showing this.

I shot and re-shot the filtered outdoor test several times with exactly the same results. It definitely wasn't controlled, but from what I remember there was no wind and the foliage in the background is a Masquite tree which doesn't move much even when the wind is blowing. I want to try again with the Tamron and see what happens (some people told me later that the 70-200 may have issues with most filters actually – thus me purchaseing a Hoya Pro MC Clear filter).

TTSkipper wrote in post #4496511
Just curious, for the first tests, you say that you sampled using the grid in PS, but did you use a single pixel to get the color or did you have the picker to select a bigger square of pixes, like 3x3 or 5x5? Since there is noise, I would think that the 3x3 or even better the 5x5 would give a more accurate assessment than a single point.

I should have been more clear – I did try to choose the same pixel but the filtered test did have more noise thus every picture I took and tried to match on the original there was a mismatch. So I did what I could and just selected a pixel right next to it that most closely matched the originals color.

Wilt wrote in post #4496773
A truly scientific test would involve taking a randomly distributed sample of shots taken with and without the filter (known only to one person helping with the experiment) and then presenting all the photos to an observer who has the challenge of identifying the filterless vs filtered photos. Then if the statistical correlation between 'identified' vs. 'real' is high enough, one can then draw the conclusion that filters do alter images to an identifiable degree.

Very good point... If someone else wanted to experiment and add thier findings here that would be great! I did shot many shots though but chose these for the samples.

Double Negative wrote in post #4497633
Filters increasing noise and affecting bokeh? I don't think so.

They might change colors slightly (if the filter has a cast, for example) or saturation (if there are reflections or flare causing loss of contrast).

The color sample size you took seems too small to me to give meaningful color values even with the grid, etc. like TTSkipper mentions.

I should have just said flare and left it as that – but included boken because that was where I noticed it.

Lester Wareham wrote in post #4498006
If you check out the Protective Filter FAQ (link in sig), you will see that others have been before you.

Comments on your results:

1. Colour Tests: Yes most UV filters will introduce a small colour bias. Look at Q16: "In the case of a Hoya SHMC Pro 1 UV(0) there was a small warming towards red of about 2 steps in 210 (red-blue) or 100K in 5250K in terms of colour temperature. ", the answer includes two small swatches. Not a big issue specially if you take a WB reference.

2. Contrast: Look at Q8: Detailed tests under controlled conditions with in frame relative light source levels about the same level as the naked sun; showed no obvious effect on contrast and for most lenses minimal additional flare and ghosting, the lens being by far the dominant factor.

An exception was deeply recessed lenses where the filter could introduce some ghosting under these very harsh conditions. My concern with your method here is being outside the conditions are not well controlled.

3. Sharpness: Look at Q8 again. This has been tested using SRF/MTF test methods, the addition of the filter made no statistically discernible difference.

Note however, that measurements to this accuracy did show significant camera AF accuracy from shot to shot with or without filter.

Unfortunately this means it is very difficult to measure for any effect on sharpness simply by just comparing two shots, the AF accuracy is just not good enough for that: For that reason this is one of the most common areas that testers have systematic test design problems.

4. Bokeh: I'm not sure what you are seeing here, perhaps you could markup a shot with arrows or circles to show what areas you are referring too. I can see some small difference in what looks like flare, but in uncontrolled conditions this could be due to movements of sun-blocking vegetation or movement of the sun itself.

Now it seems that there can be a very strange double image effects in out of focus areas when single coat filters are used (Again see Q8 for discussion but at this point no examples)

I have checked for this with Hoya SHMC Pro 1 and found no indication of this although I don't have the examples up at present. I have never heard of bokeh problems being claimed for high end filters prior to this.

Yes, everyone should read these FAQs – very informative! What I was testing was not 'filters' in general but this specific filter. I have purchased B+W multicoat filters now and will be testing them as well. To answer #4 in the last two images, look at the halo's of the OoF lights coming through the trees. In the filtered image they are larger and more predominant vs the unfiltered test. I shot more than 15 shots and they all came out the same way – I went back to the same spot to confirm this after downloading – it happened again. I think I know why now after you reminded me of this: “An exception was deeply recessed lenses where the filter could introduce some ghosting under these very harsh conditions. My concern with your method here is being outside the conditions are not well controlled.” This is what may have been causing that 'flare' like effect I discribed.

MaxxuM wrote in post #4500274
Yes, everyone should read these FAQs – very informative! What I was testing was not 'filters' in general but this specific filter. I have purchased B+W multicoat filters now and will be testing them as well. To answer #4 in the last two images, look at the halo's of the OoF lights coming through the trees. In the filtered image they are larger and more predominant vs the unfiltered test. I shot more than 15 shots and they all came out the same way – I went back to the same spot to confirm this after downloading – it happened again. I think I know why now after you reminded me of this: “An exception was deeply recessed lenses where the filter could introduce some ghosting under these very harsh conditions. My concern with your method here is being outside the conditions are not well controlled.” This is what may have been causing that 'flare' like effect I discribed.

Ok MaxxuM it's flare/ghosting and not bokeh as such, that was my guess.

Looking at this pic of the 70-200/4 IS it does not seem to be any more recessed than the 200/2.8 below, perhaps I am wrong, can you supply how many mm is between rear of the filter glass and the top of the lens front element please.

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO

HTTP response: 404 | MIME changed to 'text/html' | Byte size: ZERO

Thanks Max for the info.