fredmiranda.com Login
This topic is in the 2010 archive. You can read it and like its posts here. It is read only.

Archive 2010Sigma SD1

  
 
wingdo
Offline
Details
p.15 #1Sigma SD1
Rob Riley wrote:
will noise performance matter?

I think when a camera can offer other aspects of performance some things can be forgiven. If it can get up to an acceptable 1600 (or even 800 and a bit ) that just might hit the spot for a lot of people.


That's why I said I am looking for clean to 800 and easily cleanable to 1600 with PP. I'm not expecting D3s results here but I am expecting much better noise control than the SD15.
Mar 08, 2011 at 09:57 PM
Lotusm50
Offline
Details
p.15 #2Sigma SD1
wingdo wrote:
I'm not expecting D3s results here but I am expecting much better noise control than the SD15.



If Sigma can't achieve that, they should really pack it in.

Mar 08, 2011 at 10:34 PM
Rob Riley
Offline
Details
p.15 #3Sigma SD1
wingdo wrote:
Rob Riley wrotewill noise performance matter?

I think when a camera can offer other aspects of performance some things can be forgiven. If it can get up to an acceptable 1600 (or even 800 and a bit ) that just might hit the spot for a lot of people.
That's why I said I am looking for clean to 800 and easily cleanable to 1600 with PP. I'm not expecting D3s results here but I am expecting much better noise control than the SD15.


on e/v alone it picks up 0.43 of a stop over the 1.73x
and I guess this is the first time we will see what Sigma engineering has over the older foveon institution version which is the oldest sensor available now. I cant see it being as good as 1.52x Nikon but I dont think thats going to matter as much as some people will make it out.

In the end its about the images you take, not the BS you can walk/talk too.
Mar 08, 2011 at 11:19 PM
wingdo
Offline
Details
p.15 #4Sigma SD1
Rob I totally agree, but I do shoot low light and I am not one who can afford two completely different camera systems. I **love** the look / feel of the Foveon sensor, and I did spend a little time with a SD14. Fantastic outdoors, not good at all in low light IMO, but that is me not the camera. It really wasn't meant for low light photog.

I am hoping the SD1 changes all of that. I've always admired the look of the Foveon system.
Mar 09, 2011 at 01:02 PM
Alf Beharie
Offline
Details
p.15 #5Sigma SD1
Jorgen Udvang wrote:
Alf Beharie wrote:
I did'nt publish a conversion tutorial but basically you swap the C/Y mounting plate for a Sigma SA plate...There is no machining to do except the usual drilling securing holes...The C/Y mounting plate on the 85/1.4 is exactly the same diameter as an SA mounting plate so the SA plate fits on nicely. Then using the C'Y plate as a template you drill four new holes in the SA plate and contersink them, add the original Contax screws and your in business.


Jorgen Udvang wrote:
That sounds almost too good to be true. What about the register distance. Will that be correct as well?


Alf Beharie wrote:
It will focus to infinity no problem. I did'nt bother adding shims to get it bang on so mine could focus past infinity a little.

Jorgen Udvang wrote:
Does Sigma sell mounting plates, or will I have to cannibalize a lens? Will it work on other lenses too, like the 35-70 or the 135mm f/2.8?


Alf Beharie wrote:
Not sure without seeing them for myself.
Mar 09, 2011 at 04:22 PM
Alf Beharie
Offline
Details
p.15 #6Sigma SD1
wingdo wrote:
Rob I totally agree, but I do shoot low light and I am not one who can afford two completely different camera systems. I **love** the look / feel of the Foveon sensor, and I did spend a little time with a SD14. Fantastic outdoors, not good at all in low light IMO, but that is me not the camera. It really wasn't meant for low light photog..



Actually the SD14 is great in low light...BUT, to get good results with it in low light you MUST use low ISO (preferably ISO100 or even ISO50), a long exposure time and a tripod. Handheld nightshots with the SD14?...Forget it. Want to freeze fast moving night action?...Forget that too. Want to take low noise landscapes and architecture shots of static subjects at night? Easy peasy! Slow things down and keep it steady and you can great results:
Here is one taken with the SD14 on a very dark night indeed. The only illumination being a street light several hundred yards away to my back and the light you can see to the bottom left of frame. Lens was the Carl Zeiss 28mm Distagon T* @ f2.8:
http://img839.imageshack.us/img839/7272/cz28pano1.jpg

Mar 09, 2011 at 04:51 PM
wingdo
Offline
Details
p.15 #7Sigma SD1
Alf Beharie wrote:
Actually the SD14 is great in low light...BUT, to get good results with it in low light you MUST use low ISO (preferably ISO100 or even ISO50), a long exposure time and a tripod. Handheld nightshots with the SD14?...Forget it. Want to freeze fast moving night action?...Forget that too. Want to take low noise landscapes and architecture shots of static subjects at night? Easy peasy! Slow things down and keep it steady and you can great results:
Here is one taken with the SD14 on a very dark night indeed. The only illumination being a street light several hundred yards
...Show more →

Well that is a very different meaning of "low light" than I was thinking.

Last fall I shot a wedding where the ceremony was candlelit. My partner shoots Canon and I used my D300s mostly with a 50/1.4G (really want to buy a 1.2 AIS), and for the SD14 I used a 30/1.4 and 50/1.4 (Sigma items all borrowed and I love the feel of the 50/1.4). Due to the lighting ISO was 800 - 1200. I will have to see if we got a release for this shoot, if so I'll post a couple pictures from the SD14. None from the SD14 were used for the album. Colors were great, grain was not.

BTW, that is a great picture.
Mar 09, 2011 at 05:30 PM
Alf Beharie
Offline
Details
p.15 #8Sigma SD1
wingdo wrote:
Alf Beharie wroteActually the SD14 is great in low light...BUT, to get good results with it in low light you MUST use low ISO (preferably ISO100 or even ISO50), a long exposure time and a tripod. Handheld nightshots with the SD14?...Forget it. Want to freeze fast moving night action?...Forget that too. Want to take low noise landscapes and architecture shots of static subjects at night? Easy peasy! Slow things down and keep it steady and you can great results:
Here is one taken with the SD14 on a very dark night indeed. The only illumination being a street light several hundred yards away to my back and the light you can see to the bottom left of frame.
Well that is a very different meaning of "low light" than I was thinking.

Last fall I shot a wedding where the ceremony was candlelit. My partner shoots Canon and I used my D300s mostly with a 50/1.4G (really want to buy a 1.2 AIS), and for the SD14 I used a 30/1.4 and 50/1.4 (Sigma items all borrowed and I love the feel of the 50/1.4). Due to the lighting ISO was 800 - 1200. I will have to see if we got a release for this shoot, if so I'll post a couple pictures from the SD14. None
...Show more →

I can see your point. Candle light shots with the SD14 might be ok if you used a longer lens, say a 135mm f1.8-f2, and wide open. Then you could take close up "expression" portraits of certain individuals faces that are closest to the the light from each candle. Might be worth a try, if your ever in that situation again.

wingdo wrote:
BTW, that is a great picture.


Thanks.
Mar 09, 2011 at 06:39 PM
wingdo
Offline
Details
p.15 #9Sigma SD1
Alf Beharie wrote:
I can see your point. Candle light shots with the SD14 might be ok if you used a longer lens, say a 135mm f1.8-f2, and wide open. Then you could take close up "expression" portraits of certain individuals faces that are closest to the the light from each candle. Might be worth a try, if your ever in that situation again.

wingdo wroteBTW, that is a great picture.
Thanks.


Please do understand I am not bashing the Foveon sensor. I do like the look, it's why I was so disappointed to read the SD15 sensor is basically the one from the SD14. I think in 2011 we should expect pretty clean shots to 800 and pretty easily PP'd noise characteristics through 1600. If the SD1 cannot hit clean to 800, that means we can really only expect noise free to 400 and I don't think that is going to be a market winner unless the price for the SD1 is under 1k street.
Mar 09, 2011 at 11:01 PM
Yakim Peled
Offline
Details
p.15 #10Sigma SD1
I wonder when will it actually be shipped. Googling a bit I found a Feb 2011 estimation but B&H say that: No arrival date known.

Happy shooting,
Yakim.
Mar 10, 2011 at 04:45 AM
Alf Beharie
Offline
Details
p.15 #11Sigma SD1
wingdo wrote:
Alf Beharie wroteI can see your point. Candle light shots with the SD14 might be ok if you used a longer lens, say a 135mm f1.8-f2, and wide open. Then you could take close up "expression" portraits of certain individuals faces that are closest to the the light from each candle. Might be worth a try, if your ever in that situation again.

wingdo wroteBTW, that is a great picture.
Thanks.
Please do understand I am not bashing the Foveon sensor. I do like the look, it's why I was so disappointed to read the SD15 sensor is basically the one from the SD14. I think in 2011 we should expect pretty clean shots to 800 and pretty easily PP'd noise characteristics through 1600. If the SD1 cannot hit clean to 800, that means we can really only expect noise free to 400
and I don't think that is going to be a market winner unless the price for the SD1 is under 1k street.


Sigma claim that the SD1 has a 1-1.5 stop advantage over the SD15. They also state that it might also have ISO6400...Even if it does, I would'nt expect brilliant results from it at such a high ISO setting.
Unlike the SD14 and SD15, the SD1 will not have ISO50 capabilty, but because of its f-stop advantage Sigma claim it does'nt need it. ISO100 will therefore be the equivalent of ISO50 on the SD15 and ISO800 will therefore be the equivalent of using ISO400 on the SD15....So in theory at least, it should be a big improvement.
Personally though, I really cant fathom how they have managed to do this, being as the SD1's sensor has a somewhat higher pixel density than the current SD14/SD15 sensor.
Because the SD1's sensor is first new Foveon X3 sensor design since Sigma bought out cash strapped Foveon Inc, its possible that they may have been able to implement some neccessary hardware improvements at the photosite level (4T instead of 3T, for instance), but I suspect they have simply employed more effective low noise algorthms than in the past...But I guess we will just have to wait and see...And hope we wont have to wait too long.
As for the price...The rumour is that it will cost around the same as a Canon 7D but depending on where you get a 7D from that price could range from as little as £950 to over £2000, in the UK.
Because the SD15 was about £1000 when it was launched and the SD1 is a massive improvement, I'm guessing the actual launch price is likely to be closer to the £2000 end than the £950 end...In fact, if its released for under £1200, street, I will be very supprised indeed.
For the US price just exchange the pound (£) sign for a dollar ($) sign and it should be pretty close to the actual price there.
Mar 10, 2011 at 09:34 AM
wingdo
Offline
Details
p.15 #12Sigma SD1
Alf,

Great points and they lead right to my concerns. People talk about how moving from a 1.7 to 1.5x sensor will help with noise, but they always overlook the elephant in the room of total pixels. Indeed the SD1 will be more densely packed than the SD15. *IF* Sigma is merely employing a new low noise algorithm, why wasn't this implemented in the SD15? That would give them 2 bodies of different levels (prosumer, semi-pro / pro) with decent noise capabilities. I will continue to hold out and hope the SD1 delivers as though I'd never consider trading my Nikons for Canons, the Foveon sensor ........
Mar 10, 2011 at 10:01 AM
Rob Riley
Offline
Details
p.15 #13Sigma SD1
wingdo wrote:
Alf,

Great points and they lead right to my concerns. People talk about how moving from a 1.7 to 1.5x sensor will help with noise, but they always overlook the elephant in the room of total pixels. Indeed the SD1 will be more densely packed than the SD15. *IF* Sigma is merely employing a new low noise algorithm, why wasn't this implemented in the SD15? That would give them 2 bodies of different levels (prosumer, semi-pro / pro) with decent noise capabilities. I will continue to hold out and hope the SD1 delivers as though I'd never consider trading my Nikons
...Show more →

more pixels dont make more noise what they do make is a spatially larger image.
Reduce the size of that image down to the comparison camera size et voilà, less noise. In the case of SD14/15 and SD1 thats ROOT(16/4.7) or 1.8 stops, but 4.7Mp is pretty tiny these days perhaps the writer was thinking more conventionally.
Mar 13, 2011 at 05:00 AM
kwalsh
Offline
Details
p.15 #14Sigma SD1
Seconding Rob's comment. At these sensor sizes pixel density has nothing at all to do with image noise. Yes, it impacts pixel noise, but no one cares a hoot about that. The question is what does the noise look like in the final image. In compact cameras the pixel density has finally gotten high enough for marketing reasons that we are starting to see a degradation of IQ because the pixels are so small parasitic effects are coming into play. For APS-C sized sensors we are still a good factor of 4 away from reaching that limit. Don't worry about the pixel density...

As far as improvements go, you have to remember Foveon has only had a few iterations of its design. The SD14/SD15 sensor was really only their second generation and is quite old. There is probably quite a bit of room for improvement so it wouldn't be surprising if they achieved the 1 to 1.5 stop improvement as claimed. In contrast, the major players have a lot of design experience in their sensors and things are close to optimal so we typically don't expect big changes in IQ with each new camera.

Anyway, it will be interesting to see the results whenever the camera does come out!

Ken
Mar 13, 2011 at 07:23 AM
wingdo
Offline
Details
p.15 #15Sigma SD1
I am going to respectfully disagree regarding pixel density not affecting noise. At default ISO, yes you are right. At high ISOs I do not believe this to be correct. As you amp up the sensor sensitivity to light, it will produce more amp noise than a less dense sensor all else being equal. Diffraction also comes into play as the more dense the pixels per sq cm or sq mm, the more they become affected by diffraction.

The Nikon D3x and D3s are a great example. The D3x is a year newer than the D3s, Nikon had an extra year to work on the processor to clean up noise inherent in high ISO shooting. The D3x has twice the pixel count as the D3s, and the D3x costs 2 grand more than the D3s. All of that doesn't change the fact the D3s is far better at high ISO shooting than the D3x.

Now we have strayed FAR off topic as all I ever said was that I hope the SD1 is better at handling noise above 400ISO than the SD14/SD15 is. Apparently I've stumbled upon a thread where everyone thinks the SD15 has no noise at 1600. My apologies for not agreeing with this assumption.

Edited on Mar 19, 2011 at 11:01 PM · View previous versions
Mar 13, 2011 at 11:48 AM
Rob Riley
Offline
Details
p.15 #16Sigma SD1
nobody said any such thing
Mar 13, 2011 at 12:44 PM
kwalsh
Offline
Details
p.15 #17Sigma SD1
wingdo wrote:
I am going to respectfully disagree regarding pixel density not affecting noise. At default ISO, yes you are right. At high ISOs I do not believe this to be correct. As you amp up the sensor sensitivity to light, it will produce more amp noise than a less dense sensor all else being equal.


Actually, the opposite. "Amp noise" which would be the noise from the amplification process would be independent of pixel density and thus actually its contribution to total image noise would be lower for higher density sensors. In reality this noise source is never relevant, shot noise and kTC noise are more dominant. Photon shot noise scales with pixel density and thus final image noise contribution is density independent (again up to the point the parasitic effects and loss of fill factor become an issue). Similar results for kTC noise.


Diffraction also comes into play as the more dense the pixels per sq cm or sq mm, the more they become affected by diffraction.


Yes, there is a point at which increasing resolution is a bit senseless because of diffraction. But this has nothing to do with noise or image quality.


The Nikon D3x and D3s are a great example. The D3x is a year newer than the D3s, Nikon had an extra year to work on the processor to clean up noise inherent in high ISO shooting. The D3x has twice the pixel count as the D3s, and the D3x costs 2 grand more than the D3x. All of that doesn't change the fact the D3s is far better at high ISO shooting than the D3x.


Actually, a poor example for your point but an excellent example of the complexities involved in sensor design. The D3x actually has far superior dynamic range at lower ISO than the D3s. And the D3s has superior dynamic range at higher ISO. This has nothing to do with pixel density though, but instead with trade-offs in the sensor design. Nikon set out to achieve exactly these results. The reason for the low pixel count in the D3s is to provide for very fast readout.


Now we have strayed FAR off topic as all I ever said was that I hope the SD1 is better at handling noise above 400ISO than the SD14/SD15 is. Apparently I've stumbled upon a thread where everyone thinks the SD15 has no noise at 1600. My apologies for not agreeing with this assumption.


I didn't read anyone saying that, but I didn't parse the whole thread. If someone did make that claim I'd think they are a bit crazy!

Ken
Mar 13, 2011 at 12:53 PM
Alf Beharie
Offline
Details
p.15 #18Sigma SD1
Rob Riley wrote:
wingdo wroteAlf,

Great points and they lead right to my concerns. People talk about how moving from a 1.7 to 1.5x sensor will help with noise, but they always overlook the elephant in the room of total pixels. Indeed the SD1 will be more densely packed than the SD15. *IF* Sigma is merely employing a new low noise algorithm, why wasn't this implemented in the SD15? That would give them 2 bodies of different levels (prosumer, semi-pro / pro) with decent noise capabilities. I will continue to hold out and hope the SD1 delivers as though I'd never consider trading my Nikons for Canons, the Foveon sensor ........
more pixels dont make more noise what they do make is a spatially larger image.
Reduce the size of that image down to the comparison camera size et voilà, less noise. In the case of SD14/15 and SD1 thats ROOT(16/4.7) or 1.8 stops, but 4.7Mp is pretty tiny these days perhaps the writer was thinking more conventionally.


What wingdo meant was that the pixel density of the SD1's sensor is much higher than the SD15 sensor, even taking the SD1's 30% larger sensor size into account...This of course means that the SD1's individual photosites are smaller than those of the SD15 and therefore potentially more noisy.
He is quite correct because the larger photosites of the SD10 vs the SD14 gave the SD10 much better pixel quality than the higher res SD14, with the same sized sensor. Larger photosites do equal better pixel quality.
The extra resolution of the SD1 will help to hide some of the noise but Sigma will probably have to turn to using clever anti-noise algorithm software to hide the rest.
However, we now know that the SD1 sensor is fundimentally different down at the cicuitry level, as their latest patent shows, and this may help to reduce some of the inevitable noise:

http://www.freepatentsonline.com/y2010/0155576.html.
Mar 18, 2011 at 08:54 AM
garydavidjones
Offline
Details
p.15 #19Sigma SD1
Thanks to "thesuede" for his insightful analysis. I have

owned several Sigma SD cameras and have been alternately

frustrated and delighted with their color renditions. Love

the vibrant colors, especially photos from a trip to Vietnam

and Cambodia. On the other hand, it was maddening how

the photos often DID NOT LOOK like what I saw. Colors were

off, especially in low light situations such as museums.


In the end other problems with the Sigma SD cameras

led me to switch to Canon SLRs. Poor battery performance,

small LCD, poor performance at high ISO, etc.
Mar 18, 2011 at 09:35 AM
FM ForumsLeica & Alternative GearJoin Upload & Sell
  
 

You are not logged inLog inRegister