Our DIY Scanner
Moderator: peterZ
Re: Our DIY Scanner
The file is appended as a .CR2, so you'll have to be able to open that kind of file to view it.
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Our DIY Scanner
Attachments are temporarily set to 30mb, fire away. I'm going to look for a better system.
Re: Our DIY Scanner
Dan - I've tried twice and it doesn't seem to be working. - I tried a third time and FYI - it doesn't even come back to tell me the file is too large; maybe the "upload" got "turned off" - can you tell I don't program anything except our own little web site?
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Our DIY Scanner
Gar. I tried a 30mb file and it didn't work, either.
Re: Our DIY Scanner
And now I have to get off line; I won't be back on until sometime tomorrow. I'll gladly upload the RAW image then if it works.
- daniel_reetz
- Posts: 2812
- Joined: 03 Jun 2009, 13:56
- E-book readers owned: Used to have a PRS-500
- Number of books owned: 600
- Country: United States
- Contact:
Re: Our DIY Scanner
Until I come up with a better solution, I think drop.io is a good intermediate. You can upload the file there, and I'll host it locally after downloading it there. I PM'ed you, but it basically says the same thing I wrote here.
Re: Our DIY Scanner
Hey Dan - I dropped the file as you directed - it said that it could not convert it, which isn't a surprise; but I hope you'll be able to access it. Here's the URL:
http://drop.io/haulkgk#
http://drop.io/haulkgk#
Re: Our DIY Scanner
It's looking nice, Ann! I'm very jealous of the low amount of noise.
On the focus - I'm wondering if that's because of the extreme level of skew of the camera relative to the camera. It looks to me like it's out of focus on the side that's pointed closer to the camera; the focal length is based on the parts of the magazine that are the furthest away. I think you'll have better results if you can take your photos at as straight an angle as possible.
Looking at the histogram, it does seem that it's a bit underexposed, by about one stop - I assume that's because of the lighting environment you're using. Compared to my G10, the noise level is low enough that you can compensate for that in software without boosting the noise too much, which is very nice!
On the focus - I'm wondering if that's because of the extreme level of skew of the camera relative to the camera. It looks to me like it's out of focus on the side that's pointed closer to the camera; the focal length is based on the parts of the magazine that are the furthest away. I think you'll have better results if you can take your photos at as straight an angle as possible.
Looking at the histogram, it does seem that it's a bit underexposed, by about one stop - I assume that's because of the lighting environment you're using. Compared to my G10, the noise level is low enough that you can compensate for that in software without boosting the noise too much, which is very nice!
The opinions expressed in this post are my own and do not necessarily represent those of the Canadian Museum for Human Rights.
Re: Our DIY Scanner
Thanks Misty - I agree about the focus and the skew. I can move the "focus box" around the page before I take the photo and check it on the laptop screen and spend all day getting every bit of it in focus and never get to take a photo since one thing will be out of focus when another part will be in focus. I will never scan a magazine, but for this test of a glossy mag cover, I think the Canon XSi is fine.
As for skew, with my camera mount, there are ways to decrease the skew when taking the photo and I'm trying different things - including, if the book is small enough, putting it on one side of the cradle, scanning two pages at a time, and positioning the camera for less skew. The small example project I'm working on right now has very little skew and the focus is just fine.
I'll admit to being ignorant about how to read the histogram - it really means nothing to me. What stop do you think I should be on? I can change the camera settings through the EOS Utility. Thanks for your comments!
As for skew, with my camera mount, there are ways to decrease the skew when taking the photo and I'm trying different things - including, if the book is small enough, putting it on one side of the cradle, scanning two pages at a time, and positioning the camera for less skew. The small example project I'm working on right now has very little skew and the focus is just fine.
I'll admit to being ignorant about how to read the histogram - it really means nothing to me. What stop do you think I should be on? I can change the camera settings through the EOS Utility. Thanks for your comments!
Re: Our DIY Scanner
The remote live view in EOS Utility should allow you to use a live histogram, though I'm not quite sure how you would enable that if it's not on by default.
As an example, here's the histogram I got for your image in Photoshop, with no corrections applied:
EOS Utility's histogram should look fairly similar, though it might be slightly different
The histogram basically represents the brightness range in your image. The left edge represents blacks, while the right edge represents whites. Do you notice that empty area to the right? What that means, basically, is that your camera didn't capture any brightness in that range. The result is the image looking too dark, like you probably noticed it did straight out of the camera. For comparison, take a look at the histogram after I applied an exposure compensation of +1 in Adobe Camera Raw:
It's essentially the same shape, but basically reaches the right side of the histogram. This histogram represents a more appropriately brighter version of the image.
What you want to try to achieve is a histogram that approaches the right side side without actually touching it. Anything touching the sides causes colour clipping - which means it clips out the darkest/lightest tones. Those colours are unrecoverable if it happens, so you want to avoid it happening. It's good to leave yourself a little wiggle room. You can correct for a bit of underexposure in software, though keep in mind that it's always best to shoot a properly exposed photo in the first place rather than having to apply large corrections in software.
As for how much exposure you need to fix it - it's hard to say, but I would try +1 since the +1 exposure compensation in ACR was perfect. You can experiment with it using EOS Utility and see what gives you a balanced histogram.
As an example, here's the histogram I got for your image in Photoshop, with no corrections applied:
EOS Utility's histogram should look fairly similar, though it might be slightly different
The histogram basically represents the brightness range in your image. The left edge represents blacks, while the right edge represents whites. Do you notice that empty area to the right? What that means, basically, is that your camera didn't capture any brightness in that range. The result is the image looking too dark, like you probably noticed it did straight out of the camera. For comparison, take a look at the histogram after I applied an exposure compensation of +1 in Adobe Camera Raw:
It's essentially the same shape, but basically reaches the right side of the histogram. This histogram represents a more appropriately brighter version of the image.
What you want to try to achieve is a histogram that approaches the right side side without actually touching it. Anything touching the sides causes colour clipping - which means it clips out the darkest/lightest tones. Those colours are unrecoverable if it happens, so you want to avoid it happening. It's good to leave yourself a little wiggle room. You can correct for a bit of underexposure in software, though keep in mind that it's always best to shoot a properly exposed photo in the first place rather than having to apply large corrections in software.
As for how much exposure you need to fix it - it's hard to say, but I would try +1 since the +1 exposure compensation in ACR was perfect. You can experiment with it using EOS Utility and see what gives you a balanced histogram.
The opinions expressed in this post are my own and do not necessarily represent those of the Canadian Museum for Human Rights.