DxO produces awful results around 400-800 ISO. It greatly improves results 1600-6400. The comparison is neutral, with some benefits to DxO, at 100-320 ISO.
Personally, I'm using ACR CS4 for everything except 3200-6400 where DxO makes big difference.
David
Link to noise comparison A500
Forum rules
No more than three images or three external links allowed in any post or reply. Please trim quotations and do not include images in quotes unless essential.
No more than three images or three external links allowed in any post or reply. Please trim quotations and do not include images in quotes unless essential.
-
- Site Admin
- Posts: 5985
- Joined: Sat May 19, 2007 1:14 pm
- Location: Kelso, Scotland
- Contact:
Re: Link to noise comparison A500
David Kilpatrick wrote:DxO produces awful results around 400-800 ISO. It greatly improves results 1600-6400. The comparison is neutral, with some benefits to DxO, at 100-320 ISO.
Personally, I'm using ACR CS4 for everything except 3200-6400 where DxO makes big difference.
David
David, I'm at a loss here, can you please give some examples of "DxO produces awful results around 400-800 ISO" I don't know what to look for when comparing ACR to DxO. or are the differance between them something that would only affect Pro shooters? Also would the same go for ACR CS3?
-
- Site Admin
- Posts: 5985
- Joined: Sat May 19, 2007 1:14 pm
- Location: Kelso, Scotland
- Contact:
Re: Link to noise comparison A500
Left to its own devices DxO creates a ragged, watercolour-looking file at medium high ISO settings where ACR does this at very high ISO settings. For some reason, DxO gives a better result at higher ISOs relative to ACR. The colour conversion of DxO is generally better than CS3 but with the latest camera profiles, CS4 is producing much better colours. And it does depend on the camera body. DxO is very slow to use, and I am under pressure right now with other urgent work stuff.
But, here goes, before I start on other things:
http://www.pbase.com/davidkilpatrick/image/117164998
This shows how the controls of ACR are great for adjusting tones - dodging the foreground, altering colours, burning in the sky with a saturation mask, blurring the sky to limit grain etc; while the controls of DxO (without tonal adjustments) allow straightening of verticals with keystone correction.
http://www.pbase.com/davidkilpatrick/image/117164997
This is a close-up at 100% of an Alamy size (17 megapixel 5120 pixel high export) from both, with the DxO lightened in post-processing to reveal detail similar to ACR (which was a bit overdone anway for emphasis). While the DxO image does have higher levels of detail due to sharpening which can not be removed, and the ACR image does have visibly larger noise grains even with zero Sharpness, the DxO image has comparatively high levels of conversion artefacts creating a ragged or posterized watercolour look to edges - the ACR result is technically less 'processed'. Now you may prefer DxO even at the 640 ISO speed, but I prefer ACR. The situation changes at higher ISOs than 800, and the effects are of course not easy to spot at ISO 100-200 with either processor.
I have also put the full size 5120 pixel conversions (photo by Shirley Kilpatrick on 18-250mm and A700 by the way) on pBase, just follow to the next images.
In short, both converters are worth having. If I had an entire set of ISO 100 architectural interiors on a body and lens in the DxO database, which had been perfectly exposed (ideally, bracketed and the best selected for each) DxO would things ACR can't touch. But for a mixed bag of raws needing tonal corrections and local shading, ACR wins.
David
But, here goes, before I start on other things:
http://www.pbase.com/davidkilpatrick/image/117164998
This shows how the controls of ACR are great for adjusting tones - dodging the foreground, altering colours, burning in the sky with a saturation mask, blurring the sky to limit grain etc; while the controls of DxO (without tonal adjustments) allow straightening of verticals with keystone correction.
http://www.pbase.com/davidkilpatrick/image/117164997
This is a close-up at 100% of an Alamy size (17 megapixel 5120 pixel high export) from both, with the DxO lightened in post-processing to reveal detail similar to ACR (which was a bit overdone anway for emphasis). While the DxO image does have higher levels of detail due to sharpening which can not be removed, and the ACR image does have visibly larger noise grains even with zero Sharpness, the DxO image has comparatively high levels of conversion artefacts creating a ragged or posterized watercolour look to edges - the ACR result is technically less 'processed'. Now you may prefer DxO even at the 640 ISO speed, but I prefer ACR. The situation changes at higher ISOs than 800, and the effects are of course not easy to spot at ISO 100-200 with either processor.
I have also put the full size 5120 pixel conversions (photo by Shirley Kilpatrick on 18-250mm and A700 by the way) on pBase, just follow to the next images.
In short, both converters are worth having. If I had an entire set of ISO 100 architectural interiors on a body and lens in the DxO database, which had been perfectly exposed (ideally, bracketed and the best selected for each) DxO would things ACR can't touch. But for a mixed bag of raws needing tonal corrections and local shading, ACR wins.
David
Re: Link to noise comparison A500
Thanks David, looking at those images I could see the jagged edges in the DxO conversion and some areas areas looked better in the DXO version and "more" areas looked better in the ACR. buy the way is that a gun?
-
- Site Admin
- Posts: 5985
- Joined: Sat May 19, 2007 1:14 pm
- Location: Kelso, Scotland
- Contact:
Re: Link to noise comparison A500
I don't think it's a gun in Venice. Wear a handgun in Europe (anywhere) and you would be arrested very quickly. Just a fold in the shirt, I think!
David
David
Re: Link to noise comparison A500
Hi,
I'm not sure how closely related to this thread since it's about the A550 noise levels at 12800 but gpr 2020 had a chance to try the camera and shot sme JPGs with noise reduction set to standard and here's what he's got:
The thread is at DPreview by the way. ( it certainly isn't the worst noise I personally saw in Alphas, in fact I kind of think it's promising )
Yildiz
http://forums.dpreview.com/forums/read. ... e=33013966
I'm not sure how closely related to this thread since it's about the A550 noise levels at 12800 but gpr 2020 had a chance to try the camera and shot sme JPGs with noise reduction set to standard and here's what he's got:
The thread is at DPreview by the way. ( it certainly isn't the worst noise I personally saw in Alphas, in fact I kind of think it's promising )
Yildiz
http://forums.dpreview.com/forums/read. ... e=33013966
- Dr. Harout
- Subsuming Vortex of Brilliance
- Posts: 5662
- Joined: Wed May 30, 2007 7:38 pm
- Location: Yerevan, Armenia
- Contact:
Re: Link to noise comparison A500
At 12800 ISO that's not bad at all, considering present standards.
Re: Link to noise comparison A500
Yes, for ISO 12800 that looks very good to me too! It is downsized though. Don't know what the full size version looks like.
Bakubo http://www.bakubo.com
Re: Link to noise comparison A500
I don't need to see the full version -- it already looks as good as the ISO 1600 from my A100!bakubo wrote:Yes, for ISO 12800 that looks very good to me too! It is downsized though. Don't know what the full size version looks like.
Who is online
Users browsing this forum: No registered users and 21 guests