Ad
Ad
Ad
Pages: [1] 2 3 ... 9 »   Bottom of Page
Print
Author Topic: expose to the right?  (Read 37779 times)
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2416



WWW
« on: November 03, 2007, 04:00:57 PM »
ReplyReply

Maybe this is a dumb question....but just curious what everyone else is doing...

Seemed like expose to the right was the rule with DSLRs but since shooting my Rollei/P20 and Leica DMR  (which has similar dynamic range and is also 16bit)  I have found that I am getting better results exposing the images for the look I want rather than exposing to the right and pushing down in post.  In fact I might be crazy but to me it seems my files look best if I expose for the final rather than push down, and sometimes I find if I have to make exposure adjustments in post that the files look better if I push up rather than push down.  Am I crazy?
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
Morgan_Moore
Sr. Member
****
Offline Offline

Posts: 2218


WWW
« Reply #1 on: November 03, 2007, 04:11:01 PM »
ReplyReply

Quote
Maybe this is a dumb question....but just curious what everyone else is doing...

Seemed like expose to the right was the rule with DSLRs but since shooting my Rollei/P20 and Leica DMR  (which has similar dynamic range and is also 16bit)  I have found that I am getting better results exposing the images for the look I want rather than exposing to the right and pushing down in post.  In fact I might be crazy but to me it seems my files look best if I expose for the final rather than push down, and sometimes I find if I have to make exposure adjustments in post that the files look better if I push up rather than push down.  Am I crazy?
[a href=\"index.php?act=findpost&pid=150443\"][{POST_SNAPBACK}][/a]

Yes.

Being more serious I use a sti>dng>silkypix workflow which I don t find too good at moving the 'exposure compensation' to the left

So I always give my self a bit of 'left headroom' (footroom?) by never exppsing at 25ISO but keeping at 50 but I still go as far right as possible with the camera at 50

So my anwer is keep right, but not too far right

S
Logged

Sam Morgan Moore Cornwall
www.sammorganmoore.com -photography
jonstewart
Sr. Member
****
Offline Offline

Posts: 435


« Reply #2 on: November 03, 2007, 04:16:19 PM »
ReplyReply

Quote
Maybe this is a dumb question....but just curious what everyone else is doing...

Seemed like expose to the right was the rule with DSLRs but since shooting my Rollei/P20 and Leica DMR  (which has similar dynamic range and is also 16bit)  I have found that I am getting better results exposing the images for the look I want rather than exposing to the right and pushing down in post.  In fact I might be crazy but to me it seems my files look best if I expose for the final rather than push down, and sometimes I find if I have to make exposure adjustments in post that the files look better if I push up rather than push down.  Am I crazy?
[a href=\"index.php?act=findpost&pid=150443\"][{POST_SNAPBACK}][/a]

As a relative newbie to MFDB, I'm also finding that exposing to the right doesn't work particularly well. I'm finding outdoors that exposing down 1/2 to 1 stop (using a spot meter average) seems to be getting it just right. With lower contrast subjects, that could be different, I suppose.
Logged

Jon Stewart

If only life were so simple...
Morgan_Moore
Sr. Member
****
Offline Offline

Posts: 2218


WWW
« Reply #3 on: November 03, 2007, 04:24:21 PM »
ReplyReply

Quote
using a spot meter average)[a href=\"index.php?act=findpost&pid=150447\"][{POST_SNAPBACK}][/a]

The meter ? so 20th C  

I am looking at the histogram all the time and trying to just stop the highlight clipping (in a typical backlit scene)

S
Logged

Sam Morgan Moore Cornwall
www.sammorganmoore.com -photography
AndreNapier
Sr. Member
****
Offline Offline

Posts: 422


WWW
« Reply #4 on: November 03, 2007, 04:44:52 PM »
ReplyReply

If I trusted my Sekonic meter with A75+ I would have to redo all my sessions.
To me the amount of over or under exposure depends on the subject of the shoot.
With female faces the major issue is the quality of skin when choosing the exposure. Perfect skin
( does it really exist anymore ) I tend to expose right on or slightly under as it will work very well in the post production. Worst the skin - lighter I go and than move it to the left. It is all personal taste but it works well for me and my clients.
Andre
Logged
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2416



WWW
« Reply #5 on: November 03, 2007, 05:00:38 PM »
ReplyReply

Quote
If I trusted my Sekonic meter with A75+ I would have to redo all my sessions.
To me the amount of over or under exposure depends on the subject of the shoot.
With female faces the major issue is the quality of skin when choosing the exposure. Perfect skin
( does it really exist anymore ) I tend to expose right on or slightly under as it will work very well in the post production. Worst the skin - lighter I go and than move it to the left. It is all personal taste but it works well for me and my clients.
Andre
[a href=\"index.php?act=findpost&pid=150451\"][{POST_SNAPBACK}][/a]

Yes agree - seems to be dependent on subject - skin looks good nearly all the way to the right or even clipped a little as its the red channel that will clip first.  But things like flowers or with lots of color look better IMHO with normal exposure or even under then pushed up.
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
jonstewart
Sr. Member
****
Offline Offline

Posts: 435


« Reply #6 on: November 04, 2007, 04:51:53 AM »
ReplyReply

Quote
The meter ? so 20th C   

I am looking at the histogram all the time and trying to just stop the highlight clipping (in a typical backlit scene)

S
[a href=\"index.php?act=findpost&pid=150449\"][{POST_SNAPBACK}][/a]

See, told you I was a newbie with MF...So different from 35mm in some respects. I've now added the subject dependancy to my learning curve. Thanks guys!
« Last Edit: November 04, 2007, 04:53:04 AM by jonstewart » Logged

Jon Stewart

If only life were so simple...
rainer_v
Sr. Member
****
Offline Offline

Posts: 1131


WWW
« Reply #7 on: November 04, 2007, 06:29:20 AM »
ReplyReply

with emotion backs, together with brumbaer sw  the qestion is: where is the correct exposure point? the histogram at its right side, without showing clipped-reds in the highlight-warning, means that no channel exceeds 255.
now you still can expose one or two stops more, because the "highlight recovery" code already will find enough info to restore the image details.

this is true as long no "critical" colors have to be restored. so if i see red-warnings eg. in clouds, windows or highlights, i dont care and overexpose. if e.g. skin tones are redmarked i expose more carefull.
in general the data information in the higher stops of the histogram is much more dense than in the middle and in the shadows, for the nature how is light translated to digital.
and because digital are just many yes or nos, the image will look better as more to the right you go- ofcourse only as long there is valid information in the file.
« Last Edit: November 13, 2007, 01:23:30 PM by rainer_v » Logged

rainer viertlböck
architecture photographer
munich / germany

www.tangential.de
etrump
Full Member
***
Offline Offline

Posts: 125



« Reply #8 on: November 04, 2007, 10:55:15 PM »
ReplyReply

Quote
Maybe this is a dumb question....but just curious what everyone else is doing...

Seemed like expose to the right was the rule with DSLRs but since shooting my Rollei/P20 and Leica DMR  (which has similar dynamic range and is also 16bit)  I have found that I am getting better results exposing the images for the look I want rather than exposing to the right and pushing down in post.  In fact I might be crazy but to me it seems my files look best if I expose for the final rather than push down, and sometimes I find if I have to make exposure adjustments in post that the files look better if I push up rather than push down.  Am I crazy?
[a href=\"index.php?act=findpost&pid=150443\"][{POST_SNAPBACK}][/a]

I have found the same thing with my P30+.  I think it is because their is so much more shadow detail than 12 bit DSLRs.  I still expose to the right but don't push it as far as I used to with my Canons.  I generally expose one with a normal looking histogram and another one exposed right and 9 out of 10 times I end up using the normal exposure with a little push in the shadows.  

It is amazing how much more usable the MFDB files are in the shadows.  I have similar results with the 14 bit Canon 1DIII but to a lesser extant than the p30.
Logged

Ed Cooley

Gallery
Blog
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2416



WWW
« Reply #9 on: November 07, 2007, 03:43:38 PM »
ReplyReply

Thanks to all for your responses.

I've done some more testing with this.  I think I still prefer to not expose to the the right for most images.   Here's what I've found with my tests and I did find differences in the files exposed to the right vs properly exposed or even under exposed.  The difference has to do with where the contrast of the image is upper half or lower half.

If you expose to the right, then you will obviously need to do an extra step in post. The images from my P20/Rollei needed both a downward exposure and a gamma/brightness adjustment, but still the color saturation was different.   This put another way, if you expose right, then adjust you will decrease contrast in the upper half of the tones and increase contrast in the lower half of the tones.  If you underexpose and push up you get the opposite, increase in contrast in the brighter half of the image, and decrease in contrast in the lower half.

Comparing a file captured to expose to the right and proper (diff was .5 stops), I found the resulting processed files were not the same even if you adjust exposure, brightness and set the white and black points to the same figures!


As long as I am don't have the shadows totally blocked (or am shooting with ISO 400 or higher) then I am finding I prefer proper or even slightly underexposed images.  Call me crazy but I like the results.  

Eric
« Last Edit: November 07, 2007, 03:45:32 PM by EricWHiss » Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
digitaldog
Sr. Member
****
Offline Offline

Posts: 9166



WWW
« Reply #10 on: November 07, 2007, 04:46:59 PM »
ReplyReply

Quote
Seemed like expose to the right was the rule with DSLRs but since shooting my Rollei/P20 and Leica DMR  (which has similar dynamic range and is also 16bit)
[{POST_SNAPBACK}][/a]

Are they capturing linear encoded Raw data? If so, the ETTR concept is no different. Its not the chip size nor the bit depth at work here, its where you put your data in a linear encoded data source.

You may notice less a difference by putting less data in the shadows by not ETTR but the results are, ETTR isn't about over exposure, its about exposing to provide the most data in the last stop.

[a href=\"http://www.digitalphotopro.com/tech/exposing-for-raw.html]http://www.digitalphotopro.com/tech/exposing-for-raw.html[/url]
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2416



WWW
« Reply #11 on: November 07, 2007, 05:13:27 PM »
ReplyReply

Quote
Are they capturing linear encoded Raw data? If so, the ETTR concept is no different. Its not the chip size nor the bit depth at work here, its where you put your data in a linear encoded data source.

You may notice less a difference by putting less data in the shadows by not ETTR but the results are, ETTR isn't about over exposure, its about exposing to provide the most data in the last stop.

Andrew,
Since I am familiar with your style of posts, let me suggest you do a real test on your own. Take your MFDB and shoot a couple images on a tripod bracketing exposures and then adjust them in post so that the white points and black points are the same then adjust the brightness or gamma until the middle peaks are at the same point in the histogram. Now overlay the images in PS and set blend mode to difference.  Check it out and report back. The images will not be the same. What you have no MFBD?  okay use your favorite digicam, but I encourage you to do your own testing rather than speculate.

Eric
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
digitaldog
Sr. Member
****
Offline Offline

Posts: 9166



WWW
« Reply #12 on: November 07, 2007, 05:20:42 PM »
ReplyReply

Quote
Check it out and report back. The images will not be the same. What you have no MFBD?  okay use your favorite digicam, but I encourage you to do your own testing rather than speculate.

[a href=\"index.php?act=findpost&pid=151176\"][{POST_SNAPBACK}][/a]

I wouldn't expect them to be the same.

Again, if we're talking about linear encoded data, the encoding can't be anything other than half of all the data being contained in the first stop, the least in the last.

As for the tests, they were done with a DSLR in the URL I posted previously.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2416



WWW
« Reply #13 on: November 07, 2007, 05:26:21 PM »
ReplyReply

Quote
I wouldn't expect them to be the same.

Again, if we're talking about linear encoded data, the encoding can't be anything other than half of all the data being contained in the first stop, the least in the last.

As for the tests, they were done with a DSLR in the URL I posted previously.
[a href=\"index.php?act=findpost&pid=151177\"][{POST_SNAPBACK}][/a]


Yeah it all makes sense in a linear encoding scheme, except that the RAW converters don't provide a way to mush the data back and forth that produces equal results both for up and down. That's what I'm saying. IF there were a way to do this with C1 then ETTR would be the best way. So far I haven't found a way to do this and get equal results.  Even if you could get equal results why would you want to add one more step to your workflow if it wasn't needed? Anyhow we agree on the point that the results for normal exposure and ETTR adjusted in post are not the equal, and I am further saying I prefer the normal exposure or even .5 stop under for visual look.  I think the reason is because the contrast is not affected as it is in the ETTR adjusted version. The result looks more real to me.
« Last Edit: November 07, 2007, 05:28:41 PM by EricWHiss » Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
digitaldog
Sr. Member
****
Offline Offline

Posts: 9166



WWW
« Reply #14 on: November 07, 2007, 05:38:41 PM »
ReplyReply

Quote
Yeah it all makes sense in a linear encoding scheme, except that the RAW converters don't provide a way to mush the data back and forth that produces equal results both for up and down.

I don't understand what that means.

Quote
That's what I'm saying. IF there were a way to do this with C1 then ETTR would be the best way.

Why is C1 an issue?

Quote
So far I haven't found a way to do this and get equal results.  Even if you could get equal results why would you want to add one more step to your workflow if it wasn't needed?

They are not supposed to be equal.

Just make the highlight clipping value of each the same, let everything else fall where it will. No, the midtones and shadows will not be equal. For one, you'll see a lot less noise in the shadows (if you clamp/clip the black the same, you may end up reducing noise at the expense of true shadow detail. In fact, a great way to reduce noise is to clip the black!).


Quote
Anyhow we agree on the point that the results for normal exposure and ETTR adjusted in post are not the equal, and I am further saying I prefer the normal exposure or even .5 stop under for visual look. 

They are not. But you need to normalize the rendering for what appears to be 'over exposure' unless as I point out in the article, you really DO over expose and get past the point of sensor saturation.

Quote
I think the reason is because the contrast is not affected as it is in the ETTR adjusted version. The result looks more real to me.
[a href=\"index.php?act=findpost&pid=151180\"][{POST_SNAPBACK}][/a]

Again, I can't comment on this (looks more real). One should be able to use the Raw rendering controls to get the best possible data AND a rendering you desire but I can't speak to C1, I used Lightroom (which is the same as CR).

You'll also note, I found almost as many issues with ETTR that make it an iffy proposition. But, the math is undeniable. IF you expose properly for digital, which IS ETTR (not blowing out highlight data you hope to reproduce), you WILL get better data, less noise in that last stop. That may not be important to you and that's fine.

But the format is immaterial here if indeed we're taking apples to apples with linear encoded data. Look at the figure of this in the article, its clear how much you have in the first stop and the last stop.
Logged

Andrew Rodney
Author “Color Management for Photographers”
http://digitaldog.net/
EricWHiss
Sr. Member
****
Offline Offline

Posts: 2416



WWW
« Reply #15 on: November 07, 2007, 10:50:52 PM »
ReplyReply

Andrew,
Sorry you assumed that I read your article. I did not and most likely will not.  I think it a safe bet that other participants in this thread also will not read it.  

We agree that the data IS different. We disagree that ETTR is better data, because what matters to me is how the image looks.  

The whole point of the thread is to ask people who are shooting MFDB (which I think does not include you) whether they like the results of ETTR or not.  I am stating that I do not in most cases (though it does seem dependent on subject).  I think the reason is because it throws the contrast of the image off unequally between the lower half and upper half of the histogram.   I wanted to hear from more seasoned MFDB shooters how they felt on the issue.  

btw - I don't use the profile you made for me for my HP printer. The stock HP profile gave more realistic looking greens, yellows and reds.

Eric
Logged

Authorized Rolleiflex Dealer:
Find product information, download user manuals, or purchase online - Rolleiflex USA
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #16 on: November 08, 2007, 03:12:16 AM »
ReplyReply

Quote
Andrew,
Sorry you assumed that I read your article. I did not and most likely will not.  I think it a safe bet that other participants in this thread also will not read it.

Andrew happens to be one of the foremost exerts out there in the area of color management and digital imaging. Ignoring his input is probably unwise.
Logged

Wayne Fox
Sr. Member
****
Offline Offline

Posts: 2859



WWW
« Reply #17 on: November 08, 2007, 04:40:57 AM »
ReplyReply

Quote
Andrew,
Sorry you assumed that I read your article. I did not and most likely will not.  I think it a safe bet that other participants in this thread also will not read it. 

We agree that the data IS different. We disagree that ETTR is better data, because what matters to me is how the image looks.   

The whole point of the thread is to ask people who are shooting MFDB (which I think does not include you) whether they like the results of ETTR or not.  I am stating that I do not in most cases (though it does seem dependent on subject).  I think the reason is because it throws the contrast of the image off unequally between the lower half and upper half of the histogram.   I wanted to hear from more seasoned MFDB shooters how they felt on the issue. 

btw - I don't use the profile you made for me for my HP printer. The stock HP profile gave more realistic looking greens, yellows and reds.

Eric
[a href=\"index.php?act=findpost&pid=151239\"][{POST_SNAPBACK}][/a]

I shoot the P45+, 5d, and 1DSMarkII, and find ETTR provides far better data and less noise in shadows.  While it takes more work in the RAW processor, the end results I feel are superior - especially in the P45 files. In fact, recently I've been experimenting with a technique to simulate the  0 noise effect discussed elsewhere on this forum, and take 2 exposures, pushing all of the shadows into the upper half of the histogram on one allowing the highlights to clip.  I pull this exposure back down to match the density of the original capture, then have been trying various techniques to use this layer in the shadows.  If done right, the only change when I enable this layer is less noise .. no color or density changes.  This isn't HDR ... my dynamic range fits easily within the histogram, it's just when I'm worried about noise in shadow regions (higher ISO).

It seems to me that as long as as the sensor is capturing the data in a linear fashion, the only difference in ETTR is the ability to capture more levels/details vs noise.  As long as you don't blow pixels, the data can be pulled back down. It doesn't matter where on the histogram they lie, because the information that's important is the relationship of the values to each other, and that relationship doesn't change as long as you contain the values within the histogram - they just slide up and down equally.

I'm not sure why you wouldn't read his article, or the one by Michael.  They are both very good, well written, explain the concept very well, and discuss the role Thomas Knoll played in the concept.  At this point your evaluation of ETTR seems to be quite subjective, not based on actual science which supports the concept quite easily.
Logged

alexjones
Jr. Member
**
Offline Offline

Posts: 80


WWW
« Reply #18 on: November 08, 2007, 06:31:55 AM »
ReplyReply

I'm not sure it would be wise to write off the advice of Thomas Knoll, Andrew Rodney, Bruce Frasier and a host of others who know a great deal more about the subject than you do.  While they may not shoot for a living they did have a huge hand in creating the software that drives the industry.  With more books and articles under their belts than most can absorb, it would be better to stand on their shoulders and enjoy the view then to dismiss them so easily from you one limited view.  Just a thought.

Make a good exposure and don't blow the highlights.  In that exposure put something there that is worth looking at and don't sweat the miniscule stuff just trust that it works.
Logged
bjanes
Sr. Member
****
Offline Offline

Posts: 2794



« Reply #19 on: November 08, 2007, 07:58:26 AM »
ReplyReply

Quote
Yeah it all makes sense in a linear encoding scheme, except that the RAW converters don't provide a way to mush the data back and forth that produces equal results both for up and down. That's what I'm saying. IF there were a way to do this with C1 then ETTR would be the best way. So far I haven't found a way to do this and get equal results.  Even if you could get equal results why would you want to add one more step to your workflow if it wasn't needed? Anyhow we agree on the point that the results for normal exposure and ETTR adjusted in post are not the equal, and I am further saying I prefer the normal exposure or even .5 stop under for visual look.  I think the reason is because the contrast is not affected as it is in the ETTR adjusted version. The result looks more real to me.
[{POST_SNAPBACK}][/a]

As explained in an essay by[a href=\"http://www.luminous-landscape.com/pdf/Curves2.pdf] Mark Segal[/url] on this site, ACR curves may increase contrast and saturation, but were designed to do so by Thomas Knoll. Indeed, Mr. Knoll is reported to have said that such behavior entails extra programming effort, but he the took the extra effort because that was what most photographers want. In the case of s simple exposure correction on linear data, one would think that no such shifts would occur as Andrew Rodney stated. Exposure boost simply involves multiplication of the linear data by a given factor; for example, an exposure boost of 1 EV entails multiplying the data points by 2. All data points are affected proportionally.

Andrew is quite well known in imaging circles and the fact that you apparently do not even know who he is indicates that you are not well read and may not be that knowledgeable in imaging theory.

Please look at Mark's table and note the HSB values before and after an exposure boost of 1.75 EV. Hue and saturation do not change, only the brightness. Why don't you post some examples of the deleterious effects of a negative exposure in ACR or C1, rather than merely making undocumented claims which contradict those expressed by authorities such as Andrew, Michael Reichman, Thomas Knoll, and Bruce Fraser? Personally, I would take the view expressed by these authorities rather than the view of someone I don't know. Perhaps then I would take your claims more seriously.
Logged
Pages: [1] 2 3 ... 9 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad