Ad
Ad
Ad
Pages: « 1 [2] 3 4 ... 6 »   Bottom of Page
Print
Author Topic: Striations in P45+@800 ISO normal ?  (Read 21452 times)
eronald
Sr. Member
****
Offline Offline

Posts: 4008



WWW
« Reply #20 on: November 21, 2007, 08:26:40 PM »
ReplyReply

I am not much having much luck with the local Phase One agent Prophot, so far. They did not contact me or my dealer to say what was happening. Calling them elicited a blank, then a callback where they claimed that their sole technician did send my files on to Phase One - this tech is now on vacation.

Edmund

Quote
I have seen my dealer and given him example images. He is sending reports to Phase One. We will see how they answer.

Edmund
[a href=\"index.php?act=findpost&pid=153206\"][{POST_SNAPBACK}][/a]
« Last Edit: November 21, 2007, 08:36:00 PM by eronald » Logged

Edmund Ronald, Ph.D. 
Henry Goh
Sr. Member
****
Offline Offline

Posts: 574


« Reply #21 on: November 21, 2007, 08:42:31 PM »
ReplyReply

Quote
I am not much having much luck with the local Phase One agent Prophot, so far. They did not contact me or my dealer to say what was happening. Calling them elicited a blank, then a callback where they claimed that their sole technician did send my files on to Phase One - this tech is now on vacation.

Edmund
[a href=\"index.php?act=findpost&pid=154823\"][{POST_SNAPBACK}][/a]

Is after-sales-service a sore point with Phaseone for owners of P backs?  I seem to get that impression when reading on Phase forum.
Logged
DavidP
Full Member
***
Offline Offline

Posts: 155


WWW
« Reply #22 on: November 21, 2007, 10:26:06 PM »
ReplyReply

I think Phase One just insists that you go through a dealer, If you have a responsive dealer you won't have any trouble. In the US anyway, you have your choice of different ones to work with.
I have been very happy with mine, and I have had to deal with some issues too.

I have tried underexposing at 800 on my 45+ and bringing the exposure compensation up. The results are not that great, lots of grain and magenta green stuff in the shadows. I don't however see the kind of striation or lines that Edmund is getting.
Logged
JeffVo
Newbie
*
Offline Offline

Posts: 34


« Reply #23 on: November 21, 2007, 10:26:26 PM »
ReplyReply

I thought it common knowledge that under exposing and pushing caused this effect.  Personally I Find the Dalsa chips to look even worse.  I'd love to the see the unpushed histogram, and I'd bet its all lumped to the left.  I'll also, bet that a PROPERLY exposed shot at this iso will look fine (if it looks like this send it back). Even properly exposed images at hi iso will have "clean" hi-light and mid-tones with slightly less pleasing shadows.  Personally, I find Phase handles this situations better than most.  Underexposing any digital device and pushing always looks bad no matter how much you pay.
Logged
eronald
Sr. Member
****
Offline Offline

Posts: 4008



WWW
« Reply #24 on: November 21, 2007, 10:59:19 PM »
ReplyReply

For further entertainment, here is another ISO 800 pushed image.
Not only does this one show more stripes the more you look, but it also falls apart nicely into a left half and a right half (look at the bottom to spot the break).

I'd be very happy with this back's Hi_ISO performance if it didn't show these artefacts.

Edmund

Quote
I thought it common knowledge that under exposing and pushing caused this effect.  Personally I Find the Dalsa chips to look even worse.  I'd love to the see the unpushed histogram, and I'd bet its all lumped to the left.  I'll also, bet that a PROPERLY exposed shot at this iso will look fine (if it looks like this send it back). Even properly exposed images at hi iso will have "clean" hi-light and mid-tones with slightly less pleasing shadows.  Personally, I find Phase handles this situations better than most.  Underexposing any digital device and pushing always looks bad no matter how much you pay.
[a href=\"index.php?act=findpost&pid=154854\"][{POST_SNAPBACK}][/a]
« Last Edit: November 21, 2007, 11:04:04 PM by eronald » Logged

Edmund Ronald, Ph.D. 
vjbelle
Full Member
***
Offline Offline

Posts: 203


« Reply #25 on: November 22, 2007, 03:54:29 PM »
ReplyReply

Quote
what a great upgrade eh?
[a href=\"index.php?act=findpost&pid=153029\"][{POST_SNAPBACK}][/a]
I never did upgrade my P45....... never saw the value.  I agree with your analysis, but I still am a loyal customer of phase.  

Victor
Logged
JDG
Full Member
***
Offline Offline

Posts: 103


« Reply #26 on: November 22, 2007, 08:42:42 PM »
ReplyReply

Edmund, you could have contacted phase directly with your files and then gone to your dealer if they told you you needed to send the back in.  They were pretty quick to get back to me, with my experience dealing with my dealer (calumet) i definitely got things done quicker this way.
Logged
josayeruk
Guest
« Reply #27 on: November 23, 2007, 12:36:26 PM »
ReplyReply

Quote
For further entertainment, here is another ISO 800 pushed image.
Not only does this one show more stripes the more you look, but it also falls apart nicely into a left half and a right half (look at the bottom to spot the break).

I'd be very happy with this back's Hi_ISO performance if it didn't show these artefacts.

Edmund
[a href=\"index.php?act=findpost&pid=154859\"][{POST_SNAPBACK}][/a]

Even without the striations (is that the right word?   ) I cannot see the usefullness of this image as there is zero zilch zip shadow detail???

Jo S. x
Logged
samuel_js
Guest
« Reply #28 on: November 23, 2007, 05:21:08 PM »
ReplyReply

Quote
For further entertainment, here is another ISO 800 pushed image.
Not only does this one show more stripes the more you look, but it also falls apart nicely into a left half and a right half (look at the bottom to spot the break).

I'd be very happy with this back's Hi_ISO performance if it didn't show these artefacts.

Edmund
[a href=\"index.php?act=findpost&pid=154859\"][{POST_SNAPBACK}][/a]

Pushing the shadows so much won't give you any satisfactory result, specially if the shadows are just pure black. Could you post a sample of the original exposure? The image is sharp so I suppose you're getting really dark images aren't you? How about a flash?
« Last Edit: November 23, 2007, 05:23:26 PM by samuel_js » Logged
John Sheehy
Sr. Member
****
Offline Offline

Posts: 838


« Reply #29 on: November 23, 2007, 06:24:11 PM »
ReplyReply

Quote
Based on some conversations I have had with industry insiders, I am starting to wonder whether this is not a signal processing issue linked to the speed at which the CCD is clocked in the new P+ backs. I don't think it's really a sensor issue. Pure conjecture of course.[a href=\"index.php?act=findpost&pid=153206\"][{POST_SNAPBACK}][/a]

It depends on how you define "sensor".  It is definitely not an issue with the photosites themselves; they seem quite capable of providing a usable image here.  These are offsets in the blackpoint of entire lines from the two sensors.  That means that is has something to do with the reading and digitization of entire lines from each of the two sensors in the camera, and is therefore an issue in the readout electronics at the edge of the sensor or somewhere down the chain in the camera.  It is especially noticeable with Tungsten white balance, because the blue channel is usually multiplied by 3x to 5x for very warm light sources, while green is anchored and red is already very close to being balanced in the RAW state for warm lighting.  Where you see blue bands, the blackpoint that the camera has for that line is too low, IOW true black might not be 0, but what registers as 0.8 or 1.7 for that line (and the discrepancy is further magnified by the white balance process).
Logged
John Sheehy
Sr. Member
****
Offline Offline

Posts: 838


« Reply #30 on: November 23, 2007, 06:27:09 PM »
ReplyReply

Quote
Even without the striations (is that the right word?   ) I cannot see the usefullness of this image as there is zero zilch zip shadow detail???
[a href=\"index.php?act=findpost&pid=155250\"][{POST_SNAPBACK}][/a]

Maybe you'd see subtle detail if there weren't striations?
Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8883


« Reply #31 on: November 23, 2007, 09:40:12 PM »
ReplyReply

I can't understand what all the fuss is about. Edmund has posted an ISO 800 image on another current thread, 'High ISO versus Underexposure' at http://luminous-landscape.com/forum/index....pic=20997&st=40 where it is demonstrated that the image showing banding and striations in the shadows, as well as a lot of noise, is in fact drastically underexposed.

The only uncertainty in my mind is the degree to which the image is underexposed. My estimate from ACR is 1.7 stops. Others seem to think it's as much as 4 stops. Whether it's 4 stops underexposed or only 1.7 stops underexposed, I would say an additional 1.7 stops of exposure would have been sufficient to remove that banding and a lot of the noise.

If you want a film-like texture, wouldn't it be better to expose correctly in the first place and then add grain later in Photoshop.

Banding, striations and heavy noise are always a consequence of underexposure when the scene has a fairly high dynamic range, whatever the ISO, in my experience at least.

If I take a shot of a sunset with my Canon 5D at ISO 100 and neglect to give a full exposure to the right, I'll likely get very ugly banding in the deepest shadows.
Logged
eronald
Sr. Member
****
Offline Offline

Posts: 4008



WWW
« Reply #32 on: November 24, 2007, 06:44:25 AM »
ReplyReply

Here's a link to a Raw file EXPOSED TO THE RIGHT at 400 ISO, in DAYLIGHT.

http://download.yousendit.com/D9E8BE575BC4D993

Push it through C1 3.77 with no adjustments and you will be able to see a light line just under the stem of the pear, leading to the right. The streaking I complain about is present but less noticeable with normal exposure, but a fashion person, or color consultant will see it immediately.

Edmund
Logged

Edmund Ronald, Ph.D. 
Dustbak
Sr. Member
****
Online Online

Posts: 2368


« Reply #33 on: November 24, 2007, 07:47:54 AM »
ReplyReply

I just downloaded the file but by no means it has been exposed to the right. In ACR it can be pushed up anywhere between 1 and 2 stops before it gets acceptable. It would have needed almost 2 stops more to be exposed to the right.

Indeed I see magenta and green Gremlins especially in the dark parts but nothing that is really upsetting considering on how much I had to adjust exposure.
« Last Edit: November 24, 2007, 07:53:03 AM by Dustbak » Logged
Ray
Sr. Member
****
Offline Offline

Posts: 8883


« Reply #34 on: November 24, 2007, 09:38:04 AM »
ReplyReply

Edmund,
What on earth are you doing? Another image that's underexposed even by my very conservative standards. Here's the ACR window showing a +1.75 stop correction.

[attachment=4007:attachment]
Logged
godtfred
Sr. Member
****
Offline Offline

Posts: 293



WWW
« Reply #35 on: November 24, 2007, 10:54:43 AM »
ReplyReply

Quote
Edmund,
What on earth are you doing? Another image that's underexposed even by my very conservative standards. Here's the ACR window showing a +1.75 stop correction.

[attachment=4007:attachment]
[a href=\"index.php?act=findpost&pid=155473\"][{POST_SNAPBACK}][/a]
Quote
I just downloaded the file but by no means it has been exposed to the right. In ACR it can be pushed up anywhere between 1 and 2 stops before it gets acceptable. It would have needed almost 2 stops more to be exposed to the right.

Indeed I see magenta and green Gremlins especially in the dark parts but nothing that is really upsetting considering on how much I had to adjust exposure.
[a href=\"index.php?act=findpost&pid=155450\"][{POST_SNAPBACK}][/a]

Exposing to the right in ACR and C1 are two different things. ACR will consistently show 1-2 stops less exposure in my experience. If you run the files through C1 the histogram will be very different.

If you change the default curve in C1 from film standard to Linear response, the result is akin to ACR but will not look good... I'm a newbie to C1, so I don't know how other users deal with this.

The image is exposed too low anyway, as ACR is correct and C1 is not... so you are both right.

-axel
« Last Edit: November 24, 2007, 11:09:21 AM by godtfred » Logged

Axel Bauer
godtfred.com
H2|M679CS|P45+
Ray
Sr. Member
****
Offline Offline

Posts: 8883


« Reply #36 on: November 24, 2007, 11:32:53 AM »
ReplyReply

Quote
Exposing to the right in ACR and C1 are two different things. ACR will consistently show 1-2 stops less exposure in my experience. If you run the files through C1 the histogram will be very different.

If you change the default curve in C1 from film standard to Linear response, the result is akin to ACR but will not look good... I'm a newbie to C1, so I don't know how other users deal with this.

The image is exposed too low anyway, as ACR is correct and C1 is not... so you are both right.

-axel
[a href=\"index.php?act=findpost&pid=155498\"][{POST_SNAPBACK}][/a]

So, is this contretemp really all about differences in the way C1 and ACR handle images? A one and 2/3rds stop difference to me is so significant I find it difficult to believe that any major program like C1 could operate like that.

Does the ACR image not look good? There are hundreds of adjustments in ACR you can make if you think the image is not quite right.
Logged
Jonathan Wienke
Sr. Member
****
Offline Offline

Posts: 5759



WWW
« Reply #37 on: November 24, 2007, 11:54:18 AM »
ReplyReply

When evaluating exposure, one should always have the RAW converter's TRC set to linear. Otherwise it's too easy to misjudge whether or not you've exposed correctly when looking at the image in the RAW converter. Once exposure is set correctly, then tweak the TRC as needed/desired. This is probably the root of the controversy over whether expose to the right works with MFDBs.
Logged

Morgan_Moore
Sr. Member
****
Offline Offline

Posts: 2217


WWW
« Reply #38 on: November 24, 2007, 12:02:30 PM »
ReplyReply

Quote
These lines shouldn't be there, sure an incorrectly exposed file will have certain noise problems but lines, I dont think so. When you pay 30 grand for something you deserve to have it work correctly. And by the way, I heard from a little birdy in one of the major phase one dealers that strange lines on high iso files with the plus backs is a known issue. I'd demand a refund or a back that didnt have the problem, the non plus backs dont have this issue I heard, mine certainly doesnt, what a great upgrade eh?
[a href=\"index.php?act=findpost&pid=153029\"][{POST_SNAPBACK}][/a]

I have never owned phase but would guess that this is just horrible underxposure

whithout seeing the under unprocessed levels it is really impossible to judge

it is fundamentally a 100ISO or 50ISO chip and the fact that is works at 400 or well exposed 800 makes it a fantastic tool and a leap from the generatation of backs that crap out at about 100

It will never be a D3 or whatever

Choose the right tool for the job ??

S
Logged

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

Posts: 8883


« Reply #39 on: November 24, 2007, 12:04:11 PM »
ReplyReply

Quote
When evaluating exposure, one should always have the RAW converter's TRC set to linear. Otherwise it's too easy to misjudge whether or not you've exposed correctly when looking at the image in the RAW converter. Once exposure is set correctly, then tweak the TRC as needed/desired. This is probably the root of the controversy over whether expose to the right works with MFDBs.
[a href=\"index.php?act=findpost&pid=155514\"][{POST_SNAPBACK}][/a]

TRC? Are you referring to this?

[attachment=4013:attachment]
Logged
Pages: « 1 [2] 3 4 ... 6 »   Top of Page
Print
Jump to:  

Ad
Ad
Ad