Camera/Lens performance compared to Varicam and Panny 400.

Michael Pappas · January 9th, 2006, 07:09 PM

Here is a basic grading pass done to the H1 part to see how close it can be made to match that other camera. BTW, was that not taken from the Nick and Shannon test. If so that was not a good test. Angle was off and etc. We need to do a much better test. Had there been the 16x manual this test would look better I believe. The german mag that reviewed the H1 said 800 lines. I'm not going to dispute the Germans. Just to get a license to drive is hellish in Germany. So there very accurate. Funny though, Shannons and Nicks test had the same results. Hmm the odds, I guess it is 800 lines+.

Link to basic grading pass to match the H1 up

Quote:

Originally Posted by Matthew Greene

Ok, here's an image of the H1 split with the HDC20 (a discontinued panasonic 2.2 Mpixel 1080i camera).

This is by no means a scientific comparison, it's just a reference. Both images originated at 1920x1080 but the lens defects are clear even though they are zoomed in jpegs.

Of course this isn't a comparison of image quality, but with the other image as a comparison the lens issues should become apparent to those who haven't detected them.

The image, http://www.darrenromeo.com/assets/index.htm

Matthew Greene · January 9th, 2006, 07:15 PM

Thanks Michael, Yes, that was from the Nick and Shannon test, I'd love to see other tests being done specifically with other lenses, I too have a feeling that the 16x Manual lens wil render better images than the H1's 20x zoom. As I said I'm really pleased with what I'm seeing with the camera section and really dissapointed with the lens, do you have any ideas?

Michael Pappas · January 9th, 2006, 07:52 PM

Do you own the camera? If so:
You need to try another 20x.. When I buy SLR lenses, I always try two or three of the same model. Batch to batch can vary. The same goes for the 20x. More so now that the XL series is HD.

If your just asking about options, than I would say 14x manual or 16x manual. I am waiting for head only versions to hit the market if I go with the the H1.....

pappas

Quote:

Originally Posted by Matthew Greene

Thanks Michael, Yes, that was from the Nick and Shannon test, I'd love to see other tests being done specifically with other lenses, I too have a feeling that the 16x Manual lens wil render better images than the H1's 20x zoom. As I said I'm really pleased with what I'm seeing with the camera section and really dissapointed with the lens, do you have any ideas?

A. J. deLange · January 9th, 2006, 07:57 PM

I assume the one on the left is the H1? It shows about 1 pixel's CA whereas the one on the right shows about 10.

Matthew Greene · January 9th, 2006, 08:37 PM

Quote:

Originally Posted by A. J. deLange

I assume the one on the left is the H1? It shows about 1 pixel's CA whereas the one on the right shows about 10.

The one on the right is... From Shannon's test. It's zoomed in so you won't be able to judge how many pixels the abberation covers. It's pretty bad but we don't really know what stop it was shot at and Michael also says it wasn't properly shot. If that's with the iris wide open I think I'd be fine with that but if it's stopped down a bit I need another lens.

Quote:

Originally Posted by Michael Pappas

Do you own the camera? If so:
You need to try another 20x.. When I buy SLR lenses, I always try two or three of the same model. Batch to batch can vary. The same goes for the 20x. More so now that the XL series is HD.

If your just asking about options, than I would say 14x manual or 16x manual. I am waiting for head only versions to hit the market if I go with the the H1.....

pappas

I don't own it yet, I'm waiting for it to be available without the 20x lens since I have no use for it. I have no need for it until April, hopefully by then it will be an option to just get the body, I also rather wait till it's no longer at list price since they will most likely drop the price once all the early adopters buy.

I'm seriously considering the manual lenses as well for many reasons but I'd like to see some tests before I commit to them. I'd have no problem investing on a good (expensive) general purpose lens for the H1 if that were an option.

A. J. deLange · January 11th, 2006, 08:27 AM

After all this staring at images I figured I'd try actually measuring the lateral CA of the stock H1 lens which I did using one of the edges on an ISO 12233 chart (printed on the office printer I grant you) which filled the image out at about 1 meter (accurate framing isn't important here I'm just using the edge). The point spread functions for red and blue were separated by 0.67 pixel and the edge was about 872 pixels out from the center (near the edge of the frame). This yields a CA of approximately 0.08% which is right on the border line between "Low. Not visible unless you look for it" and "Moderate. Somewhat visible at high print magnification" (according to Imatest, an outfit that makes software for measuring CA. Note: I did not use their software.)

So, with the usual caveats that this is an anecdotal result and my chart wasn't too good and I didn't do gamma correction and I didn't follow 12233 and etc it looks at first blush as if this lens really isn't so bad after all.

Petr Marusek · January 11th, 2006, 10:36 AM

That means very low CA. You will have CA on every camera with pixel shift.

Matthew Greene · January 11th, 2006, 12:51 PM

But at what stop? The CA observations are pointless unless you specify what aperture and focal length it was calculated at. Keep in mind that it can go from horrid at wide open iris to invisible at a higher stop. I'd like to see some charts shot at different apertures.

Quote:

Originally Posted by Petr Marusek

That means very low CA. You will have CA on every camera with pixel shift.

Once again, it's not the color fringing that bothers me the most from the lens abberations, it's the loss of resolving power, contrast in the lens. That's not caused by pixel shifting, in fact, pixel shifting doesn't cause visible fringing either, just a slight shadowing around pixels at most.

Petr Marusek · January 11th, 2006, 01:00 PM

Quote:

Originally Posted by Matthew Greene

it's not the color fringing that bothers me the most from the lens abberations, it's the loss of resolving power, contrast in the lens.

The lens resolves over 800 lines and has excellent contrast. Why are you so down on this Canon camera? You have provided misleading information from the very first post.

Matthew Greene · January 11th, 2006, 01:27 PM

I love the camera, I hate the lens, if you remember, my initial question was "what options are there to the lens?".

I can't see why anyone would say it's a "great" lens, I mean, how realistic is that?

I haven't provided any real information, just observations that I clearly stated were not scientific in nature. I simply looked at images shot with the camera and the first thing I saw was fuzzy borders, low corner contrast and fringing.

I'm not putting the camera down, I just want a better lens. Now I ask you, what's the point of trying to convince everyone that this is a "great" lens when every real world example says otherwise. I mean, anyone with an eye and a basic understanding of optics can see through that statement right away.

Why can't anyone answer what fstop the images are at? I've said before that if the issues only show up at extreme ends of the scale I don't mind it.

I don't want hype, just real world information. For a camera that has interchangeable lens I want to pick a lens that works for me. That's it, is it too much to ask?

A. J. deLange · January 11th, 2006, 01:51 PM

This test was done at f/2 with the camera a meter from an 8x10 chart. Clearly I'd like to do CA and MTF comparisons between the 3x, 16x, stock and some Nikon lenses at 30f and 60i and though I don't personally use 24f in that mode as well. And it should be done at all apertures and focal lengths. Clearly this would mean defining a new career for myself so it's going to have to be a sampling for the time being at least. I want to polish this software some before going into "production" but also want to get bits out to people. The reports of good pictures keep coming in and the measured data seems to support the notion that the lens is pretty good - at least thus far. There may well be some disappointments ahead but we'll talk about those when we hit them.

I don't want to say too much about MTF until I get a good test chart but I will say that the crude chart shows about 20% response at 350 cy/ph (roughly equivalent to 700 lpph). Four hundred cy (800lines) would be a bit of a stretch as the response is lower still.

Matthew Greene · January 11th, 2006, 01:56 PM

Great, now you're talking! Pick a medium focal length and expose the MTF chart at every stop. I'd go with 60i just because it's the native format. This would give finally give good data on the lens.

Michael Pappas · January 11th, 2006, 05:49 PM

Simple solution,

Buy the manual lens. If I get the H1, I'm getting the manual lens. You have two options 16x and a 14x. The 16x is a current product and is a very good piece of glass. All tests have shown that the 16x is better than the 20x. Ok now that obstacle is out of your way...

Quote:

Originally Posted by Matthew Greene

I love the camera, I hate the lens, if you remember, my initial question was "what options are there to the lens?".

I can't se?

Matthew Greene · January 11th, 2006, 06:16 PM

I'm sorry this was turned into a war between fanatics and investigators but I want real world results, not hype. Yes, the manual lens shots I've seen look better and along with the 3x, it's my current option. Just have to wait for the camera head to sell alone.

I am hoping for something better than a $1300 lens at some point though. When you get to consider the physics of optics some serious engineering needs to be put into a proper lens for a 1/3" HD sensor. Fujinon's 13x is probable closer to the idea of a properly engineered HD lens, if there only was a way to test it with the H1. The pinnacle is just not going to be a lens originally designed for an SD camera, or maybe it will, guess we've got to wait on Canon.

Petr Marusek · January 11th, 2006, 06:49 PM

Look at it this way. The camera resolves about the same as the Varicam and the Varicam has $30,000 lens, in addition to the 65,000 body. If you're so unhappy with the Canon performance and you will be waiting for Canon to sell you the camera, instead for $9,000, for $8,000 without a lens, do so, but quit badmouthing the Canon, telling BS fairy tales how dealers give 20-30% discount on Varicam, telling us about the Panasonic $32K model 400 performance, and when you get caught that it is not true, then telling us about some guy from South America that has a resolution chart. You started with BS and your BS does not stop. What do you really want besides doing everything you can to discredit the Canon?

January 9th, 2006, 07:15 PM	#17
Matthew Greene Regular Crew Join Date: Jan 2006 Location: Las Vegas, NV / Branson, MO Posts: 63	Thanks Michael, Yes, that was from the Nick and Shannon test, I'd love to see other tests being done specifically with other lenses, I too have a feeling that the 16x Manual lens wil render better images than the H1's 20x zoom. As I said I'm really pleased with what I'm seeing with the camera section and really dissapointed with the lens, do you have any ideas?

January 9th, 2006, 07:57 PM	#19
A. J. deLange Major Player Join Date: Jun 2004 Location: McLean, VA United States Posts: 749	I assume the one on the left is the H1? It shows about 1 pixel's CA whereas the one on the right shows about 10.

January 11th, 2006, 08:27 AM	#21
A. J. deLange Major Player Join Date: Jun 2004 Location: McLean, VA United States Posts: 749	After all this staring at images I figured I'd try actually measuring the lateral CA of the stock H1 lens which I did using one of the edges on an ISO 12233 chart (printed on the office printer I grant you) which filled the image out at about 1 meter (accurate framing isn't important here I'm just using the edge). The point spread functions for red and blue were separated by 0.67 pixel and the edge was about 872 pixels out from the center (near the edge of the frame). This yields a CA of approximately 0.08% which is right on the border line between "Low. Not visible unless you look for it" and "Moderate. Somewhat visible at high print magnification" (according to Imatest, an outfit that makes software for measuring CA. Note: I did not use their software.) So, with the usual caveats that this is an anecdotal result and my chart wasn't too good and I didn't do gamma correction and I didn't follow 12233 and etc it looks at first blush as if this lens really isn't so bad after all.

January 11th, 2006, 10:36 AM	#22
Petr Marusek Major Player Join Date: Dec 2005 Posts: 155	That means very low CA. You will have CA on every camera with pixel shift.

January 11th, 2006, 01:27 PM	#25
Matthew Greene Regular Crew Join Date: Jan 2006 Location: Las Vegas, NV / Branson, MO Posts: 63	I love the camera, I hate the lens, if you remember, my initial question was "what options are there to the lens?". I can't see why anyone would say it's a "great" lens, I mean, how realistic is that? I haven't provided any real information, just observations that I clearly stated were not scientific in nature. I simply looked at images shot with the camera and the first thing I saw was fuzzy borders, low corner contrast and fringing. I'm not putting the camera down, I just want a better lens. Now I ask you, what's the point of trying to convince everyone that this is a "great" lens when every real world example says otherwise. I mean, anyone with an eye and a basic understanding of optics can see through that statement right away. Why can't anyone answer what fstop the images are at? I've said before that if the issues only show up at extreme ends of the scale I don't mind it. I don't want hype, just real world information. For a camera that has interchangeable lens I want to pick a lens that works for me. That's it, is it too much to ask?

January 11th, 2006, 01:51 PM	#26
A. J. deLange Major Player Join Date: Jun 2004 Location: McLean, VA United States Posts: 749	This test was done at f/2 with the camera a meter from an 8x10 chart. Clearly I'd like to do CA and MTF comparisons between the 3x, 16x, stock and some Nikon lenses at 30f and 60i and though I don't personally use 24f in that mode as well. And it should be done at all apertures and focal lengths. Clearly this would mean defining a new career for myself so it's going to have to be a sampling for the time being at least. I want to polish this software some before going into "production" but also want to get bits out to people. The reports of good pictures keep coming in and the measured data seems to support the notion that the lens is pretty good - at least thus far. There may well be some disappointments ahead but we'll talk about those when we hit them. I don't want to say too much about MTF until I get a good test chart but I will say that the crude chart shows about 20% response at 350 cy/ph (roughly equivalent to 700 lpph). Four hundred cy (800lines) would be a bit of a stretch as the response is lower still.

January 11th, 2006, 01:56 PM	#27
Matthew Greene Regular Crew Join Date: Jan 2006 Location: Las Vegas, NV / Branson, MO Posts: 63	Great, now you're talking! Pick a medium focal length and expose the MTF chart at every stop. I'd go with 60i just because it's the native format. This would give finally give good data on the lens.

January 11th, 2006, 06:16 PM	#29
Matthew Greene Regular Crew Join Date: Jan 2006 Location: Las Vegas, NV / Branson, MO Posts: 63	I'm sorry this was turned into a war between fanatics and investigators but I want real world results, not hype. Yes, the manual lens shots I've seen look better and along with the 3x, it's my current option. Just have to wait for the camera head to sell alone. I am hoping for something better than a $1300 lens at some point though. When you get to consider the physics of optics some serious engineering needs to be put into a proper lens for a 1/3" HD sensor. Fujinon's 13x is probable closer to the idea of a properly engineered HD lens, if there only was a way to test it with the H1. The pinnacle is just not going to be a lens originally designed for an SD camera, or maybe it will, guess we've got to wait on Canon.

January 11th, 2006, 06:49 PM	#30
Petr Marusek Major Player Join Date: Dec 2005 Posts: 155	Look at it this way. The camera resolves about the same as the Varicam and the Varicam has $30,000 lens, in addition to the 65,000 body. If you're so unhappy with the Canon performance and you will be waiting for Canon to sell you the camera, instead for $9,000, for $8,000 without a lens, do so, but quit badmouthing the Canon, telling BS fairy tales how dealers give 20-30% discount on Varicam, telling us about the Panasonic $32K model 400 performance, and when you get caught that it is not true, then telling us about some guy from South America that has a resolution chart. You started with BS and your BS does not stop. What do you really want besides doing everything you can to discredit the Canon?