Digital zoom, what are the facts? Some people say that it's merely an internal crop of the image on the ccd. Other people say that there is some in-cameral processing that improves the quality of the final image. Which is correct?
-
Welcome to the new forums! Please read this first. For known issues we are working to resolve, click here.
Digital Zoom, a piece of crop???
- Thread starter Frash
- Start date
Both might be correct in some cases. You can take a picture without digital zoom and then enlarge it in software and do all sorts of tricks to improve it. I am not convinced that in camera processing is going to be better. It would take too long to do anything serious and processing power is at a premium in a digicam.
vk
vk
Petteri Sulonen
Forum Pro
It is an internal crop and resample of the image. The resampling method depends on the camera; some are better than others, but in no case it will be better than what you can get in Photoshop, let alone specialized software like Genuine Fractals. Forget digital zoom.
Petteri
--
http://homepage.mac.com/psulonen/
bluh
Forum Enthusiast
Here's what I did a while ago to compare. The digital zoom is very soft but this is at normal sharpness setting, not -1, so I don't know how much better it might have been. But then both ways were using the same setup, JPG format. These crops are from the 2272x1704 originals (wind was blowing, btw, but fast enough shutter). The important thing here, I believe, is that digital zoom could be a quick way to get a closeup photo from lack of telephoto lens converter. So I don't dismiss it entirely myself. If I remember right, the optical crop here was bicubic interpolated in PS otherwise it would have been very pixelated; the digital is untouched. Got to keep in mind though that this is all jpeg though.
I think someone needs to prove a purely optical image can be cropped to match the 11 X digital zoom and still be clean and sharp. I have only tried what I can and nothing looks great to me so far. Canon G2's DZ is only too soft not real bad, in my opinion.
--
Farewell, Bob H.
Gus Stangeland
Senior Member
I hear this said a lot, but two things, first not very many G2 owners have Genuine Fractals or would want to invest the time and money. Second, in my opinion, the proof is in the pudding. I have seen VERY few real-world examples comparing digital zoom with software "zoom". To me, that's really all that matters - academically saying it's "no good" seems to me like "throwing the baby out with the wash". I would recommend that people considering taking a photo with digital zoom, take one with zoom and another of the same subject without digital zoom. Then when you get back home you can always, at any time, use software to see if you can get a better image. I'm NOT saying the software image won't be better, I'm just saying as a practical matter, use the digital zoom if you want to.
Here's Delicate Arch at Arches National Park taken with full zoom on the G2, a 2X Tiffen telephoto plus 2.7X digital zoom for a total 35mm equivalent of 550mm. Tripod, manual focus set to infinity, self-timer. Those of you that have been there know that it is a very difficult 3 mile walk to get up the the arch, so this photo was taken from the "viewing area" which was almost a mile away.
If anyone wants to do a "software digital zoom" to show that it is better than this, I would be happy to e-mail several shots I took without digital zoom.
Some time back someone (Nahau I believe) did a thorough investigation and found G1 DZ was in fact better than Photoshop resampling. It was marginal but in my opinion it really was better.
Don't know why. Perhaps they have a really good algorithm for resampling the RAW CCD data.
But generally, I agree; forget digital zoom. If you can't get closer, take a picture of something else.
--
---------------
http://edsphotos.us/
Gus Stangeland
Senior Member
OK, I'll respond to my own post.
Gus Stangeland wrote:
The top photo is Delicate Arch at Arches National Park taken with full zoom on the G2, a 2X Tiffen telephoto plus 2.7X digital zoom for a total 35mm equivalent of 550mm. Tripod, manual focus set to infinity, self-timer. Those of you that have been there know that it is a very difficult 3 mile walk to get up the the arch, so this photo was taken from the "viewing area" which was almost a mile away.
The lower photo is same as above but without digital zoom. I just cropped the arch out of the original photo and scaled it (down) to match the size of the top one. To me, I cannot see any difference. Most likely some of you will contiune to say the digital zoom is no good. You know, are we forgetting these are digital cameras - unless you are desperate for Compact Flash space, it costs NOTHING to take a few digital photos which require no post-processing when you get home. If you don't like them, delete them. I don't see the point in saying "don't EVER use digital zoom". But, that's just my opinion - I plan to use it whenever I feel like it.
Gus Stangeland wrote:
The top photo is Delicate Arch at Arches National Park taken with full zoom on the G2, a 2X Tiffen telephoto plus 2.7X digital zoom for a total 35mm equivalent of 550mm. Tripod, manual focus set to infinity, self-timer. Those of you that have been there know that it is a very difficult 3 mile walk to get up the the arch, so this photo was taken from the "viewing area" which was almost a mile away.
The lower photo is same as above but without digital zoom. I just cropped the arch out of the original photo and scaled it (down) to match the size of the top one. To me, I cannot see any difference. Most likely some of you will contiune to say the digital zoom is no good. You know, are we forgetting these are digital cameras - unless you are desperate for Compact Flash space, it costs NOTHING to take a few digital photos which require no post-processing when you get home. If you don't like them, delete them. I don't see the point in saying "don't EVER use digital zoom". But, that's just my opinion - I plan to use it whenever I feel like it.
mike savad
Senior Member
i like your title.
x2 resamples the normal graphic, and readjusts the settings for that new shot. the internal crop is better looking then one done in photoshop. x4 is a doubling of the x2 - and it does a terrible job. i only use it when i want to read something from very far away.
---Mike Savad
--
http://www.pbase.com/savad/
http://www.photosig.com/userphotos.php?id=9050
grablife
Senior Member
My argument for not using digital zoom has nothing to do with quality of one over the other. If you use digital zoom, you are locked into the crop you chose. It cannot be recropped. If you take the shot without digital zoom, you are free to recrop as desired in post processing. The only case I might use digital zoom is in a situation where I cannot do post processing and will be showing the photo on a tv. Even then, I would take a second photo w/o digital zoom to have some flexibility later.
Check out the web's most extensive Canon G2 vs Nikon CP5000 comparison at:
http://www.igrablife.com/digitalphoto
If you use http://www.pbase.com , please support them by making a small donation at http://www.pbase.com/support
For the last time, the Canon G2 DOES NOT need a circular polarizer. Stop wasting your money. Send it to me instead.
--
Check out the web's most extensive Canon G2 vs Nikon CP5000 comparison at:
http://www.igrablife.com/digitalphoto
If you use http://www.pbase.com , please support them by making a small donation at http://www.pbase.com/support
For the last time, the Canon G2 DOES NOT need a circular polarizer. Stop wasting your money. Send it to me instead.
mike savad
Senior Member
the one time digi zoom came in handy was when i was shooting between braches. the camera kept locking onto the tree and not the center, no matter what i tried. switching it to digi, allowed the camera to lock onto target without a problem.Some time back someone (Nahau I believe) did a thorough
investigation and found G1 DZ was in fact better than Photoshop
resampling. It was marginal but in my opinion it really was better.
Don't know why. Perhaps they have a really good algorithm for
resampling the RAW CCD data.
But generally, I agree; forget digital zoom. If you can't get
closer, take a picture of something else.
--
---------------
http://edsphotos.us/
---Mike Savad
--
http://www.pbase.com/savad/
http://www.photosig.com/userphotos.php?id=9050
I decided to do a test. These were shot with a cheap set of Kenko 1,2,4 close up lenses stacked together. The image on the left was cropped and enlarged in Photoshop. The one on the left is an 11x digital zoom.
I didn't enlarge in steps (1% or 5%) as is suggested in the Genuine Fractals review I read. This is just a one step enlargement with bicubic interpolation. Still, I think the digital zoom image looks much better. What do you think?
http://www.pbase.com/image/2493277
Aside from the image quality, it's interesting how the flash exposure seems to have changed between the two. The digital zoom image gave a better exposure. It was also easier to focus.
I didn't enlarge in steps (1% or 5%) as is suggested in the Genuine Fractals review I read. This is just a one step enlargement with bicubic interpolation. Still, I think the digital zoom image looks much better. What do you think?
http://www.pbase.com/image/2493277
Aside from the image quality, it's interesting how the flash exposure seems to have changed between the two. The digital zoom image gave a better exposure. It was also easier to focus.
Thanks for the tip on GF. That looks very effective.
bluh
Forum Enthusiast
An opposing argument I know of would be about metering the exposure correctly for the specific thing you want to zoom in on. Not that it isn't impossible to spot meter on a place, reposition view and crop. Just that again it's a matter of convenience when you could pop a CF or MD into a home printer and get easy print of what you were photographing through use of DZ. I want to try it more and learn of the true possiblities of both methods before I conclude anything myself. Again I say, I haven't seen someone work magic on a cropped optical zoom and read how it was done. Or if I have I've forgotten.
--
Farewell, Bob H.
Gus Stangeland
Senior Member
As I have said several times before - simply take 2 shots, one with DZ and the other without DZ. Then you have ALL options - and it didn't cost a cent. Why limit your options?
bluh
Forum Enthusiast
Yep, I suppose you used JPG format like I did so to compare in the same file format. You end up zooming in on the artifacts which can be both like sharpening and fuzzing at once. The DZ tends to be so much cleaner since the artifacts remain at a smaller scale. Still, as you said, even with bicubic resampling in PS on a TIF from CRW conversion it's likely to have more mosaic showing of pixels I think. But I also won't discount the possiblity of optical cropping being better until I've seen enough proof of it first. And I'm going to want to know the procedure of course ;-)
--
Farewell, Bob H.
That's correct. They were both JPEG. I didn't want to hassle with changing the menu and later converting RAW, and I don't think I'd want to in the field either. If I get really carried away I might compare RAW/TIF as well. It was just shoot, crop, and enlarge. The sample is a screenshot bitmap of both images side by side in pixel view.
I'm especially interested in using DZ for macro work, so the exposure and focusing attributes are making DZ look like a very attractive option.
Yep, these look just like the comparison images that I got. Resizing in one jump, the digital zoom image looks better. I, too, wonder how much better it would look with more sharpening, in or out of camera. Without buying and using Genuine Fractal or step-resizing in 1-5% increments, I don't think cropping can produce as good an image.
Keyboard shortcuts
- f
- Forum