Is the SL1 sensor an improvement?

[Lee Jay]

DPReview published some JPEGs from a pre-release camera. Still, I can't do this with my T2i. Here I've pushed blacks that were in the neighborhood of 8 counts up to the neighborhood of 140 counts. I see no sign of banding and very little noise.

What do you think? Better noise processing? Black point clipping? Or a genuine reduction of read noise? Raw images will settle this, but I haven't seen any yet.

Here's the animated gif:

http://tinyurl.com/dxr747l

 

[Mt Spokane Photography]

Its at least two generations newer than your T2i. Some seem to think that the number of pixels defines a sensor generation. It doesn't. This is supposedly a second generation Hybrid sensor.

As far as IQ, at low ISO's you are almost certainly not going to notice any difference. Improvements to sensors are most noticible by the reduced noise at high ISP's, or, in this case, by better hybrid autofocus.

The bottom line is that its a improvement for focusing in the live mode, and probably has a bit lower noise at high ISO'sin jpeg mode due to the new Digic 5. Certainly no big difference unless you want autofocus while taking video.

 

[Lee Jay]

As far as IQ, at low ISO's you are almost certainly not going to notice any difference.

Not true.

Canon's problem has always been very high read noise at low-ISO, limiting dynamic range. For example, the 5DIII's base-ISO read noise is 33.1e- while the D800's is 2.7e-:

http://www.sensorgen.info/CanonEOS_5D_MkIII.html
http://www.sensorgen.info/NikonD800.html

There's a huge difference between 33.1e- and 2.7e- when you have to push the shadows really hard. I don't run into base ISO dynamic range problems a lot, but I wouldn't mind Canon getting competitive here so that when I did I didn't struggle with the noise produced.

Canon's sensors have always been superior or competitive at high ISO where read noise drops dramatically and performance is dictated by Quantum Efficiency where the 5DIII and D800 are nearly the same at around 50%.

 

[hamada]

Its at least two generations newer than your T2i. Some seem to think that the number of pixels defines a sensor generation. It doesn't. This is supposedly a second generation Hybrid sensor.

second generation of a sensor that has worse results then the 550D sensor (by DxO tests).

im curious if the downward trend in DxO tests continues.
the rebel sensors don´t get much attention from canon when it comes to better IQ.

 

[Loswr]

It is true, he asked how it would compare to his T2i, not a D800. A D800 or Nikon was not mentioned by the OP!
Its not going to be much different from his T2i.

Since when has that stopped anyone from bringing up the comparison? :

 

[Skulker]

It is true, he asked how it would compare to his T2i, not a D800. A D800 or Nikon was not mentioned by the OP!
Its not going to be much different from his T2i.

Since when has that stopped anyone from bringing up the comparison? :

and when will it stop them :

 

[Lee Jay]

It is true, he asked how it would compare to his T2i, not a D800. A D800 or Nikon was not mentioned by the OP!
Its not going to be much different from his T2i.

Since when has that stopped anyone from bringing up the comparison? :

and when will it stop them :

You guys are off on an irrelevant tangent. I compared a 5DIII to a D800. But okay, how about a D7000 to a 60D?

http://www.sensorgen.info/NikonD7000.html

3.1e-

http://www.sensorgen.info/CanonEOS_60D.html

13.2e-

Still a very big difference in low-ISO read noise, which tends to drive dynamic range at low ISO.

 

[thelebaron]

You should go hang around bestbuy's photography section leejay, with some of these handy charts.

 

[pdirestajr]

I think the SL1 is super cute. They should release it in a hot pink variation. Then we can talk about how much people love color noise!

 

[Aglet]

I'm also very curious about whether any low ISO FPN has improved in this generation.
However, OOC jpg is not a great indicator, even jpgs from my bandy 7d looked good, raw held the real problem.

I can bide my time... waiting for something truly good to come along from Canon, as I already have truly good from other mfrs to use in the meantime.
Heck, i even sold my EF 70-200 f/2.8 L v2 the other day, since it doesn't fit on my d800... :-\

 

[Don Haines]

I think the SL1 is super cute. They should release it in a hot pink variation. Then we can talk about how much people love color noise!

I hear they have a version in camo...... but nobody's seen it yet

 

[Don Haines]

It has to be an improvement.... a new version of focusing on the sensor for live view... and a lot more computing power that can hopefully run a better/faster algorithm for determining focus.

I regard focusing ability as the most important aspect of a camera. I've said it before and I'll say it again, megapixels, dynamic range, ISO capability, and great glass don't matter if the shot is out of focus.

 

[pj1974]

As far as IQ, at low ISO's you are almost certainly not going to notice any difference.

Not true.

"DPReview published some JPEGs from a pre-release camera. Still, I can't do this with my T2i. Here I've pushed blacks that were in the neighborhood of 8 counts up to the neighborhood of 140 counts. I see no sign of banding and very little noise.

What do you think? Better noise processing? Black point clipping? Or a genuine reduction of read noise? Raw images will settle this, but I haven't seen any yet."
It is true, he asked how it would compare to his T2i, not a D800. A D800 or Nikon was not mentioned by the OP!Its not going to be much different from his T2i.

I appreciate the input of many on this forum... but what I think Mt Spokane and Neuro have missed in this, is that Lee Jay IS the OP. I am following Lee Jay's line of thinking here....

At the end of the day I hope there is improvement at both Low ISO and High ISO ... ie less noise, low banding, etc.

Paul

 

[jrista]

@LeeJay: What ISO is that sample image?

I can lift the shadows of my 7D at ISO 400 or greater by quite a bit and not experience banding noise. Problem is, I have already lost ANOTHER stop of DR by ISO 400 relative to ISO 100/200, so that really isn't saying much.

If the sample GIF shots are ISO 100, I have to say...that looks fairly impressive. I would have expected banding noise to show up in an ISO 100 shot lifted that much.

 

[Lee Jay]

That was an ISO 100 shot.

 

[jrista]

That was an ISO 100 shot.

Well, I guess one of two things then. Either the shot was not DR limited and the shadows were not deep enough to exhibit the classic Canon banding problems (which seems to be the more plausible), or Canon has definitely done something with that sensor to reduce banding noise. The lower-end Canon DSLRs usually have less anyway, as the ADC runs at a lower frequency...but there is usually some.

 

[Lee Jay]

That was an ISO 100 shot.

Well, I guess one of two things then. Either the shot was not DR limited and the shadows were not deep enough to exhibit the classic Canon banding problems (which seems to be the more plausible), or Canon has definitely done something with that sensor to reduce banding noise. The lower-end Canon DSLRs usually have less anyway, as the ADC runs at a lower frequency...but there is usually some.

Yes...it's going to take raw data to tell for sure, but it looks promising so far.

 

[Mt Spokane Photography]

As far as IQ, at low ISO's you are almost certainly not going to notice any difference.

Not true.

"DPReview published some JPEGs from a pre-release camera. Still, I can't do this with my T2i. Here I've pushed blacks that were in the neighborhood of 8 counts up to the neighborhood of 140 counts. I see no sign of banding and very little noise.

What do you think? Better noise processing? Black point clipping? Or a genuine reduction of read noise? Raw images will settle this, but I haven't seen any yet."
It is true, he asked how it would compare to his T2i, not a D800. A D800 or Nikon was not mentioned by the OP!Its not going to be much different from his T2i.

I appreciate the input of many on this forum... but what I think Mt Spokane and Neuro have missed in this, is that Lee Jay IS the OP. I am following Lee Jay's line of thinking here....

At the end of the day I hope there is improvement at both Low ISO and High ISO ... ie less noise, low banding, etc.

Paul

The point is that he asked if the upcoming SL-1 was going to be better than his T2i.
Then when answerwed that it probably would not be, he said "not true". And started in about the D600 which was not mentioned in the first place.

He needs to ask the question he wants answered.

 

[Lee Jay]

The point is that he asked if the upcoming SL-1 was going to be better than his T2i.

I asked no such question at all.

I asked the rhetorical question of if the posted push of an SL1 JPEG was, "Better noise processing? Black point clipping? Or a genuine reduction of read noise?" It's rhetorical because no one can answer that conclusively until we get some raw data from this camera.

 

[Loswr]

I asked no such question at all.

I asked the rhetorical question ... It's rhetorical because no one can answer that conclusively until we get some raw data from this camera.

So...it was a completely pointless thread. Well, gee, now I know to ignore your posts henceforth. Lesson learned.