Ad
Ad
Ad
Pages: « 1 2 3 [4] 5 »   Bottom of Page
Print
Author Topic: Comparisons From DT's 250 and 260 testing in the Library  (Read 6048 times)
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7406


WWW
« Reply #60 on: February 15, 2014, 03:32:48 AM »
ReplyReply

Hi,

Yes I am aware of that. But it doesn't list it as a recommended colour space.  I also know that Capture One has something they call Profoto which may or may not be RIMM, and they list all ICC profiles installed on the system. I felt that using Adobe RGB (which is one of the recommended ones) was the safer bet.

Best regards
Erik


Hi Erik,

CCapture One does have the capability to export as ProPhoto RGb (see screencapture attachment).
 


Bart
« Last Edit: February 16, 2014, 12:31:48 AM by ErikKaffehr » Logged

tho_mas
Sr. Member
****
Offline Offline

Posts: 1696


« Reply #61 on: February 15, 2014, 05:10:18 AM »
ReplyReply

I also know that Capture One has something they call Profoto which may or may not be RIMM
there's no such output profile. Output profiles contained in C1 are sRGB, AdobeRGB and 3 different Phaseone greyscale profiles. (Of course you can also convert to any icc profile installed on your system ... as you've mentioned.)

I felt that using Adobe RGB (which is one of the recommended ones) was he safer bet.
Huh
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7406


WWW
« Reply #62 on: February 15, 2014, 05:23:00 AM »
ReplyReply

Sorry,

What is your point? I used Adobe RGB, What should I use instead for proper comparison. ProPhoto RGB is not recommended.

And, I actually don't think it matters at all.

Best regards
Erik

there's no such output profile. Output profiles contained in C1 are sRGB, AdobeRGB and 3 different Phaseone greyscale profiles. (Of course you can also convert to any icc profile installed on your system ... as you've mentioned.)
 Huh
Logged

tho_mas
Sr. Member
****
Offline Offline

Posts: 1696


« Reply #63 on: February 15, 2014, 05:49:15 AM »
ReplyReply

What is your point?
1.) your statement about C1 containing an output profile that is "something they call Profoto which may or may not be RIMM" is incorrect.
2.) By default C1 only displays a "short list" of profiles (input profile embedded, AdobeRGB, sRGB). It doesn't necessarily mean that any of these icc profiles are preferred. It's just an idiot-proof list of profiles (well, it would be really idiot-proof if they would would leave out "embed input profile" of this short list... but that's a different story). When you click on "show all" C1 displays all the icc profiles installed on your system and of course C1's CMM converts to these color spaces just like any other (ICC aware) imaging software. Why would you think that AdobeRGB is a "safer bet"?
In fact, when you read C1's documentation you will find Phaseone recommends to use ProPhoto RGB.
Logged
Paul2660
Sr. Member
****
Offline Offline

Posts: 1820


WWW
« Reply #64 on: February 15, 2014, 06:05:54 AM »
ReplyReply

Hi,

This is my impression from this test:

  • The IQ-250 works as expected, it is a Sony sensor yielding the DR Sony sensor normally deliver. Works as expected.
  • Well possible that it does not perform as well CCD cameras with shift. Doug and Torger have good insights in this.
  • Much less shadow noise from the CMOS sensor than from the CCD, but that has been expected.
  • The test says little about colour rendition. My guess is that it is pretty good.
  • I also note that the camera produces a lot of moiré, colour aliasing and aliasing in general in this setting, but that is expected from combination of excellent lenses, relatively large pixels and no OLP filtering. Nice that image processing theory still works! ;-)

My take is that CMOS is probably a good thing for MFD. Personally I use live view a lot for accurate focusing. If it was not for live view I would still shoot my Sony Alpha 900 from 2009, but it is now retired.

I would say that the camera is quite expensive for 1.3X crop, but it is beyond my means anyway. Phase One obviously feels they have chosen a correct price point for the camera, that is their decision.

Best regards
Erik


Hi Eric

Thanks for the detailed color testing results were interesting.  

On your summary of the 250 you mentioned color aliasing issues,  I felt the 250 did much better then the 260 shot in regards to the color aliasing.  Do you feel the 250 still has issues here?

Thanks
Paul C
Logged

Paul Caldwell
Little Rock, Arkansas U.S.
Photography > http://photosofarkansas.com
Blog> http://paulcaldwellphotography.com
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7406


WWW
« Reply #65 on: February 15, 2014, 08:08:46 AM »
ReplyReply

Hi Paul,

Thanks for raising the issue. I may have mixed up some pictures. I will review my observations, but it may take some time.

You are probably right.

Best regards
Erik




On your summary of the 250 you mentioned color aliasing issues,  I felt the 250 did much better then the 260 shot in regards to the color aliasing.  Do you feel the 250 still has issues here?

Thanks
Paul C

Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7406


WWW
« Reply #66 on: February 15, 2014, 08:18:40 AM »
ReplyReply

Hi,

I recall a long discussion about that profile, but I cannot find it. I will correct my text on the issue.

Regarding #2, in the export dialog you three options are shown. "Embed camera profile, Adobe RGB and sRGB" in addition there is an option "show all", if you choose that, the bottom line says shows recommended, which takes you back to the original dialog with three options.

The interpretation I have made is that those profiles that are shown when click on recommended are the recommended profiles.

Best regards
Erik


1.) your statement about C1 containing an output profile that is "something they call Profoto which may or may not be RIMM" is incorrect.
2.) By default C1 only displays a "short list" of profiles (input profile embedded, AdobeRGB, sRGB). It doesn't necessarily mean that any of these icc profiles are preferred. It's just an idiot-proof list of profiles (well, it would be really idiot-proof if they would would leave out "embed input profile" of this short list... but that's a different story). When you click on "show all" C1 displays all the icc profiles installed on your system and of course C1's CMM converts to these color spaces just like any other (ICC aware) imaging software. Why would you think that AdobeRGB is a "safer bet"?
In fact, when you read C1's documentation you will find Phaseone recommends to use ProPhoto RGB.

Logged

tho_mas
Sr. Member
****
Offline Offline

Posts: 1696


« Reply #67 on: February 15, 2014, 09:58:34 AM »
ReplyReply

I recall a long discussion about that profile, but I cannot find it. I will correct my text on the issue.
the respective discussion was about an input profile, not an output profile.

Quote
The interpretation I have made is that those profiles that are shown when click on recommended are the recommended profiles.
a similar recommendation like: "use sRGB for a monitor and use AdobeRGB for print". It's a recommendation for useres completely unfamiliar with color management. Nothing more and nothing less.
Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7406


WWW
« Reply #68 on: February 15, 2014, 03:57:28 PM »
ReplyReply

OK,

That explains it. In the context it matters very little. Both LR5 and Capture One are a bit more accurate in Prophoto RGB than in Adobe RGB. But whatever colour space, LR 5 is almost twice as accurate as C1, under studio flash conditions.

Regarding the profile, I just recalled seeing the discussion and it made me a bit insecure. Anyway all the fields in the ColorChecker fit within Adobe RGB, so Adobe RGB is appropriate for the intended use.


Best regards
Erik



the respective discussion was about an input profile, not an output profile.
a similar recommendation like: "use sRGB for a monitor and use AdobeRGB for print". It's a recommendation for useres completely unfamiliar with color management. Nothing more and nothing less.
« Last Edit: February 15, 2014, 04:06:49 PM by ErikKaffehr » Logged

Paul2660
Sr. Member
****
Offline Offline

Posts: 1820


WWW
« Reply #69 on: February 15, 2014, 05:04:06 PM »
ReplyReply

I quit using Prophoto over a year ago.  The Epson 9900 can't fully print to Adobe 1998  and Prophoto is way out there.  I got tired of being out of gamut to my printer, especially on matte canvas, which seems to have the tightest gamut anyway.  
I realize that 16bit printing is possible on the MAC, but I windows based.  Also I don't that the Epson 9900 can begin to handle the full gamut of 16 prophoto color space.  I just decided to convert to a more manageable color space.

And the web still runs on sRGB pretty much.  

Paul C.

« Last Edit: February 15, 2014, 05:05:47 PM by Paul2660 » Logged

Paul Caldwell
Little Rock, Arkansas U.S.
Photography > http://photosofarkansas.com
Blog> http://paulcaldwellphotography.com
BartvanderWolf
Sr. Member
****
Offline Offline

Posts: 3643


« Reply #70 on: February 15, 2014, 05:23:59 PM »
ReplyReply

Regarding the profile, I just recalled seeing the discussion and it made me a bit insecure. Anyway all the fields in the ColorChecker fit within Adobe RGB, so Adobe RGB is appropriate for the intended use.

Hi Erik,

Which just means that the CC colors are not as saturated as some real world colors might be. Especially flowers can exhibit extremely saturated colors (e.g. Red is high, while Green (or Blue) is near zero). Perhaps (Bruce Lindbloom's) BetaRGB is a safer Colorspace than Adobe RGB, unless you know that the particular scene is sufficiently encoded with the Adobe RGB gamut.

The take home lesson is to use the smallest although large enough colorspace, to preserve small quantization steps for a large enough gamut, for one's output files.

Cheers,
Bart
Logged
eronald
Sr. Member
****
Offline Offline

Posts: 4026



« Reply #71 on: February 15, 2014, 05:26:58 PM »
ReplyReply

the respective discussion was about an input profile, not an output profile.
a similar recommendation like: "use sRGB for a monitor and use AdobeRGB for print". It's a recommendation for useres completely unfamiliar with color management. Nothing more and nothing less.

I think users completely unfamiliar with color management is a good description of most working pros without science degrees.

Frankly, setting everything in sight to sRGB and letting the color management fly itself on autopilot and autothrottle may be the smartest thing which a photographer can do.

If you go to web, sRGB is expected, if you do your own prints you can get decent color that way with the vendor drivers which are first rate these days, and if you are going to press, people downstream are responsible for what happens.

Edmund
« Last Edit: February 15, 2014, 06:23:14 PM by eronald » Logged
Fine_Art
Sr. Member
****
Offline Offline

Posts: 1087


« Reply #72 on: February 15, 2014, 07:49:57 PM »
ReplyReply

Most people are pretty comfortable with ARGB. Why not go ARGB all the way to output, then switch to sRGB if it is for monitor/web?
Logged
synn
Sr. Member
****
Offline Offline

Posts: 398



WWW
« Reply #73 on: February 15, 2014, 08:01:53 PM »
ReplyReply

I quit using Prophoto over a year ago.  The Epson 9900 can't fully print to Adobe 1998  and Prophoto is way out there.  I got tired of being out of gamut to my printer, especially on matte canvas, which seems to have the tightest gamut anyway.  
I realize that 16bit printing is possible on the MAC, but I windows based.  Also I don't that the Epson 9900 can begin to handle the full gamut of 16 prophoto color space.  I just decided to convert to a more manageable color space.

And the web still runs on sRGB pretty much.  

Paul C.



prophotoRGB is very much an intermediate color space; not an output one, kinda like the prores format is for video editing. Finish all your edits in prophotoRGB, use the convert to profile dialog box in Photoshop, chose "Perceptual" and output to whatever your preferred output colorspace is.
This has worked really well for me so far.
Logged

my portfolio: www.sandeepmurali.com
eronald
Sr. Member
****
Offline Offline

Posts: 4026



« Reply #74 on: February 15, 2014, 09:41:18 PM »
ReplyReply

Most people are pretty comfortable with ARGB. Why not go ARGB all the way to output, then switch to sRGB if it is for monitor/web?

This is a perfectly good strategy - in fact it is what I mostly do myself. However you'll have to remember to convert any image you externalize in some way. If you don't chances are that the next person in the work chain will foobar the image by losing the profile and using a workflow that assumes untagged=srgb. A typical situation where this will happen will be in legacy web browsers.

If you just adopt sRGB, you lose slightly in gamut, but everything just works. In some case it might even work better, because sRGB is basically what has been hard-coded into most of the consumer architecture, but I'm not going into this today.

Edmund
« Last Edit: February 15, 2014, 09:50:11 PM by eronald » Logged
ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7406


WWW
« Reply #75 on: February 16, 2014, 01:24:17 AM »
ReplyReply

Hi,

It depends on the output. Many printers can print outside Adobe RGB, so for printing it makes a lot of sense staying within Prophoto RGB.

For web display sRGB is the best choice, but that may change with 4K which has a much wider colour space.

In my case it is easy, as I normally work in Lightroom which uses a colour space with Prophoto RGB primaries, do everything in Prophoto RGB and export as needed.

But, this discussion started about accuracy of colour rendition. What I have seen is that exporting in Adobe RGB reduces colour accuracy in both LR 5.3 and Capture 1.

Best regards
Erik


prophotoRGB is very much an intermediate color space; not an output one, kinda like the prores format is for video editing. Finish all your edits in prophotoRGB, use the convert to profile dialog box in Photoshop, chose "Perceptual" and output to whatever your preferred output colorspace is.
This has worked really well for me so far.

Logged

ErikKaffehr
Sr. Member
****
Offline Offline

Posts: 7406


WWW
« Reply #76 on: February 16, 2014, 02:47:54 AM »
ReplyReply

Hi,

You are absolutely right, I mixed up the images! Very little colour aliasing in the IQ-250 shot!

Best regard
Erik



Hi Eric

On your summary of the 250 you mentioned color aliasing issues,  I felt the 250 did much better then the 260 shot in regards to the color aliasing.  Do you feel the 250 still has issues here?

Thanks
Paul C

Logged

gerald.d
Sr. Member
****
Offline Offline

Posts: 387


« Reply #77 on: February 18, 2014, 08:01:12 AM »
ReplyReply

Hi,

You are absolutely right, I mixed up the images! Very little colour aliasing in the IQ-250 shot!

Best regard
Erik




I'm a bit surprised there's not been more discussion about this.

The 250 absolutely blows the 260 and 280 out of the water.

What's going on here?
Logged
torger
Sr. Member
****
Offline Offline

Posts: 1486


« Reply #78 on: February 18, 2014, 08:58:00 AM »
ReplyReply

Concerning color aliasing bit I'd guess it's about the smaller pixel size of the IQ250 so you get a tad more lens blur and thus less color aliasing. I would not expect that large difference, but perhaps it by luck happened to hit some magic threshold value where the demosaicer makes a significant better job (color aliasing is much affected by how the demosaicer work).

The IQ260 would have gained from being shot at f/11 in this scene, the lens is just too sharp for the sensor Smiley

Edit: oh... it was the IQ280 not IQ260... hmm... pixel size difference is not so large there right?
Logged
gerald.d
Sr. Member
****
Offline Offline

Posts: 387


« Reply #79 on: February 18, 2014, 10:26:30 AM »
ReplyReply

Concerning color aliasing bit I'd guess it's about the smaller pixel size of the IQ250 so you get a tad more lens blur and thus less color aliasing. I would not expect that large difference, but perhaps it by luck happened to hit some magic threshold value where the demosaicer makes a significant better job (color aliasing is much affected by how the demosaicer work).

The IQ260 would have gained from being shot at f/11 in this scene, the lens is just too sharp for the sensor Smiley

Edit: oh... it was the IQ280 not IQ260... hmm... pixel size difference is not so large there right?

Yup - very little difference. IQ250 FF sensor would be 75MP IIRC.
Logged
Pages: « 1 2 3 [4] 5 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad