Sharpest Lens Forum Challenge - DPChallenge Forums

05/10/2005 11:25:05 AM · #51

-

Message edited by author 2005-05-11 15:15:03.

cpanaioti

Canon EF 70-300mm f/4.5-5.6 DO IS USM

05/10/2005 11:32:29 AM · #52

Originally posted by ovenbird:

Just a little tidbit, according to Canonâ��s Chuck Westfall, RAW images from all Canon cameras with the exception of the 1-series bodies receive some level of sharpening in the camera.

If you shoot with the sharpening turned all the way down you will get essentially the same level of sharpening as what you would get from a RAW file that has been converted with sharpening turned off.

Tom

RAW is what the sensor sees so even if anything is applied in camera can be undone in the RAW converter whereas with jpeg what's done is done.

edit: removed photo comment

Message edited by author 2005-05-10 11:39:28.

05/10/2005 11:39:04 AM · #53

-

Message edited by author 2005-05-11 15:15:31.

cpanaioti

Canon EF 70-300mm f/4.5-5.6 DO IS USM

05/10/2005 11:44:42 AM · #54

Originally posted by ovenbird:

Yes, I know it is difficult to believe, but if you have a Canon DSLR and it isnâ��t a 1-series body your RAW files are NOT exactly what the sensor recorded. Instead they have received some level of processing in the camera and it can not be undone. This is information directly from Chuck Westfall of Canon USA. If he doesnâ��t know what he is talking about (in the realm of Canon DSLRâ��s) then no one does.

100% crops are soft, that is a fact of life (for now anyway). That is of course unless you have a camera with no AA filter (Kodak 14# or Sigma SD#).

The test for the lens is how much detail is in there. In my 100% crop there is quite a bit of detail which is a pretty good showing for a 4x zoom shot wide open at an extreme focal length.

Tom

Originally posted by cpanaioti:

Originally posted by ovenbird:
Just a little tidbit, according to Canonâ��s Chuck Westfall, RAW images from all Canon cameras with the exception of the 1-series bodies receive some level of sharpening in the camera.

If you shoot with the sharpening turned all the way down you will get essentially the same level of sharpening as what you would get from a RAW file that has been converted with sharpening turned off.

Tom

RAW is what the sensor sees so even if anything is applied in camera can be undone in the RAW converter whereas with jpeg what's done is done.

Whatever.

Your image looks very sharp at websize however the 100% crop does look at bit soft.

Maybe you could post the reference.

Canon EF 16-35mm f/2.8L USM

05/10/2005 11:50:47 AM · #55

Tom, if Canon is indeed doing that they are doing the industry a disservice and pulling the wool over a lot of peoples eyes. A raw file is supposed to be unsharpened by the camera so you can apply your own choice of sharpening. You would not want a USM applied first.

Luminous Landscape, states it best:

"Because a raw file has not been processed in any way, if new and improved methods of linearizing files, applying colour filter array decoding, or other image processing advances are made, you can return to your archived raw files and work on them afresh. A JPG file, on the other hand, is fully baked."

Reference at Luminous Landscape

In any case, please post only RAW 100% crop files here per the guidelines that this thread was created under. It would be great if you could shoot a new example for us of what looks like a fine lens, if you don't already have a raw example. You might also want to try ISO 100 or 200. I'm dissapointed to see what looks like noise in an XT shot at ISO 400! I ordered the XT as an upgrade to my Rebel so I can start using ISO 400-1600 more. I haven't received it yet, and I'm having second thoughts, particularly after reading the Pop Photography review.

Message edited by author 2005-05-10 11:54:29.

Falc

Canon EOS-1D Mark III

05/10/2005 11:59:26 AM · #56

Here are examples from a 10D with 16-35 f2.8 'L' lens using 100 iso setting at the 16mm end of the zoom

The first is the full frame image

The second is the 100% crop

Nicely sharp on the stonework.

Message edited by author 2005-05-10 12:02:13.

Nikon AF-S DX NIKKOR 18-200mm f/3.5-5.6G ED VR II

05/10/2005 12:11:22 PM · #57

-

Message edited by author 2005-05-11 15:15:59.

Judith Polakoff

Nikon D7000

05/10/2005 01:22:08 PM · #58

The Nikon 50mm F/1.8 is definitely my sharpest lens:

A seashell.

Camera settings: ISO 200, 1/1000 sec., F/11

05/10/2005 01:35:14 PM · #59

-

Message edited by author 2005-05-11 15:16:43.

cpanaioti

Canon EF 70-300mm f/4.5-5.6 DO IS USM

05/10/2005 01:43:10 PM · #60

Even though Chris Westfall sticks by his story (as told to him by Canon tech support) there is still some confusion on the issue of sharpening of CRW raw files. One question raised was how can sensor data be manipulated without first interpolating it? Also, the File Viewer Utility was mentioned several times so it is unclear whether the sharpening happens in camera or when FVU reads the data from the CRW file. I don't use FVU.

Back to your (ovenbird) suggestion. It would be interesting to compare results from a jpeg with the settings you posted and a CRW file with sharpening set to zero in the converter. ACR has sharpening set to 25 as the default.

05/10/2005 03:01:34 PM · #61

-

Message edited by author 2005-05-11 15:18:01.

Tamron SP AF 28-75mm f/2.8 XR Di for Canon

05/10/2005 10:40:18 PM · #62

-

Message edited by author 2005-05-11 15:19:29.

ADGibson

Canon EOS-20D

05/11/2005 01:35:46 AM · #63

Ok. Here is my copy of John's card pic with my Tamron 28-75 2.8

And the 100% crop

I don't think it's as good as the Canon L, but it is damn close I think.

Adding settings:
Canon 20D
Tamron 28-75mm
Shutter 1/125
F8.0
ISO 100
Focal Length 75mm
RAW processor: Canon Digital Photo Professional (this is the one that came with the camera)

I think John is right. Playing card sucks for this. His recommendation was a new $1 bill. While I'm out I may try and get a couple. That may give me a better idea how sharp this lens is. Maybe a shorter focal length would give better results. I could even open it up all the way and see how it performs. If anyone is interested.

All I can say is I'm learning alot about how my cam operates in these conditions.

Message edited by author 2005-05-11 11:49:05.

kirbic

Canon EOS R5

Canon RF 24-70mm f/2.8 L IS

05/11/2005 01:42:00 AM · #64

Just an observation, but most decent lenses will produce images with resolution equaling or exceeding the resolution of the sensor when stopped down into their sweet spot. Even 100% crops are therefore not necessarily telling.
As an example, I have lots of very sharp images from my trusty old Canon 28-200 (now owned by coolhar). Though that lens is not well rated, I always found it did very nice work when operated within its envelope. The thing that separates decent lenses from great ones is performance at the edges of the envelope.

Tamron SP AF 28-75mm f/2.8 XR Di for Canon

05/11/2005 01:51:55 AM · #65

Originally posted by kirbic:

Yes, I thought about that. But I for one was interested whether a lens could produce a sharp image on its own--without processing. And also what level of blur was typical, while the sharpened photos still shine.

So that meant to me, "best shot unprocessed". There are clearly differences among the lenses shown here in the RAW set. So far, to me, the sharpest looks like the Tamron 28-75 (assuming those from DoFear were RAW, he didn't say that), and the Canon 50mm. My Sigma still looks to me to be pretty high up in this set. I don't see my 70-200mm/F4 being sharper than it. The juries still out (my own jury) on my Canon 70-300 IS DO.

ADGibson

Canon EOS-20D

05/11/2005 02:08:23 AM · #66

Mine were shot as RAW files. Sorry I should have posted that

rich

Canon EF 50mm f/1.4 USM

05/11/2005 11:04:10 AM · #67

I'd love to see this guy post some shots from this lens.

05/11/2005 11:10:40 AM · #68

-

Message edited by author 2005-05-11 15:20:02.

05/11/2005 12:26:49 PM · #69

-

Message edited by author 2005-05-11 15:20:57.

05/11/2005 12:31:36 PM · #70

Originally posted by ovenbird:

no comments? This was just starting to get interesting.

Tom

You should note that when I started this thread, I posted a cuople of caveats about using RSE as the RAW converter. I mentioned them below (or above, depending on your order) in my first post. Did you turn of Detail extraction AND sharpening? See the article I pointed to as well. It compares various converters, and points out that RSE STILL does sharpening, even when turned off completely. Most (perhaps all) converters actually can turn off sharpening.

Edit: Oops. I note that I didn't post the article reference here, I posted it in some other threads.

Here it is:

Raw Converters Review

Edit 2: I recall seeing last night, in the latest RSE, that there's a "Processing Bias Parameters" dialog under the ? menu. It says: "These bias parameters will be applied in addition to the corrections made in the correct panel and are used with all RAW images originating from the camera model specified below".

Pehaps this is why it still sharpens even though both detail and sharpening sliders are set to -50.

Message edited by author 2005-05-11 12:37:57.

Canon EF 70-200mm f/2.8L IS USM

05/11/2005 12:38:48 PM · #71

-

Message edited by author 2005-05-11 15:22:04.

theSaj

Canon EOS-20D

05/11/2005 12:59:29 PM · #72

I actually think the unconverted RAW philosophy is flawed...and here is why.

As mentioned, anti-aliasing, and other effects "soften" the results. The camera has built in compensation that accounts / counters / balances the effect of this treatment.

In truth, the question is NOT how sharp a RAW image you can get. But rather how sharp AND clear a final image you can get. A mathematician cares about raw sharpness. A photographer truly cares about the final output.

For example, if your camera took a cleaner "RAW" file but my camera's RAW file processed to an image with more clarity than your RAW file. I would definitely want the image with the most clarity (whether camera treated or not).

To restrict a cameras compensation and correction for it's post-processing (anti-aliasing/etc) just doesn't make sense to me.

*shrug*

05/11/2005 01:02:47 PM · #73

Originally posted by ovenbird:

Yes I did turn them down. I still think your request for RAW converted images is a silly way to do it as it is less standardized than just shooting jpg with a specified set of parameters. The difference wasnâ��t because of detail extraction or sharpening, it is simply contrast. The sharpening applied by the lowest sharpening parameter is the same as a RAW file converted without sharpening.

I am getting the strong impression that this thread wasnâ��t actually about learning anything. I canâ��t quite figure out what the point was at this time as you have been given some very good information but you choose to ignore it.

Tom

If I may respectfully submit: It's a bit presumptuous of you to assume I've ignored anything, nor that I haven't gotten anything out of this thread.

On the other hand, you've come in, asserted the methodology was useless, and rather then just ignoring what others seem to think was interesting as well and were participating in, you've insisted it should be done your way. Actually, your arguments and posts have been confrontational, and perhaps have been the biggest "killer" of this thread. At least that's my feeling why I sort of stopped looking forward to watching it. Perhaps others felt that too.

Your post of RAW/JPEG was interesting, but doesn't really prove anything. I am not trying to convince you to use RAW, but this was the "challenge" I proposed. Like the regular challenges, if you don't want to play, you don't have to.

Not trying to be inflammatory, but I did want to address your assertion.

Edit: I should add, that I certainly would also be interested in a separate thread/challenge to look at sharpening in camera only using JPEG. But in order to do that, you need to make sure everyone has the same parameters set. So if you'd like, go ahead and start a thread like that, propose standard parameters, and I will be happy to follow it and even participate.

Message edited by author 2005-05-11 13:07:41.

05/11/2005 01:08:53 PM · #74

-

Message edited by author 2005-05-11 15:23:00.