Pixel 2 XL vs Canon SL1/100D

Galeta · Mar 30, 2018

Hi!

While using DPreview camera studio scene comparison tool I was barely able to notice the quality difference between Pixel 2 XL and Canon SL1/100D.

Is this me who do not know where and how spot difference properly or there is not that much of it?

noisephotographer · Mar 30, 2018

When you compare jpgs, then the comparison shows that when both cameras get the same exposure, there will be slightly less noise in low light and certain areas will have slightly more detail. In low light you get approximately the same exposure when you choose Iso 800 for the Canon cameras: https://www.dpreview.com/reviews/im...1&x=-0.5143090912498635&y=-0.7628346952750105

But keep in mind that there is nearly no f/1.8 wide-angle lens for Canon cameras. When you want to compare the approximate noise performance with an f/3.5 lens, then you need to choose Iso 3200 for the Canon when you compare it with the Pixel 2.

Apart from less noise and more resolution, the Canon will take worse wide-angle jpg photos than the Pixel 2 due to worse algorithms. If you don't use a fast or stabilized lens, the noise/details could be even worse possibly.

Wayne Larmon · Mar 30, 2018

Galeta said:
Hi!

While using DPreview camera studio scene comparison tool I was barely able to notice the quality difference between Pixel 2 XL and Canon SL1/100D.

Is this me who do not know where and how spot difference properly or there is not that much of it?

It isn't you. I've been comparing my Pixel 2 against my Canon M6 for about a month. Some results were posted in this thread.

I haven't changed my opinion much since that thread. Which is that OTC Pixel 2 JPEGS are often comparable in quality to M6 raw files that were tweaked with CC 2018 ACR (described in the above-linked thread.)

What that thread doesn't cover was the rest of the pictures I took on vacation. Because it was a vacation I didn't spend a lot of time taking matching Pixel 2/M6 images. But when I was going through my images after I got home I found that M6 images weren't appreciably better than Pixel 2 images. If both cameras captured the same scene and the Pixel 2 got a shot I wanted, I went with the Pixel 2 and didn't attempt processing the M6 raw.

Mainly the only time I used M6 images were when I needed more reach (55-200mm lens) than I could get with zooming-by-cropping Pixel 2 images. Or wanted the 10mm focal length I got with my 10-22mm lens. Or the two or three images that might be wall hangers and 24 megapixels were needed.

Otherwise clean 12 megapixel Pixel 2 images more than sufficed for vacation pictures, displayed on the highest resolution monitor I have (currently my 2160x440 Surface Pro 3.) Zooming into the images indicates that Pixel 2 images would hold up at higher resolutions. (4K is about 8 megapixels...)

PM me if you want a link to these images.

Wayne

noisephotographer · Mar 30, 2018

Quite surprising that Canon 100D+f/3.5 lens would have noisier jpgs than Pixel 2 (as I said you need to choose the Iso 3200 images in order to see f/3.5 noise performance in this case): https://www.dpreview.com/reviews/im...1&x=0.07595711330975598&y=-0.1286091875974752

In theory f/3.5 on a Canon Aps-c camera is equivalent to a 1/2.55" sensor (Pixel 2) and f/0.95 lens.

Wayne Larmon · Mar 30, 2018

noisephotographer said:
Quite surprising that Canon 100D+f/3.5 lens would have noisier jpgs than Pixel 2

My experience is that they (Pixel 2 and M6) are roughly equivalent. In good light, both produce noise free images that look good at 100%. In poor light they produce noisy images that are best not looked at too closely. I haven't found a circumstance where one produces quieter images than the other.

Based on the real world pictures I've taken to date. I posted most of the ad hoc test images I have in my Pixel 2 vs. M6 thread.

DPReview has indicated that comparing cameras like the Pixel 2 to DSLRs using the DPReview test scene is problematic.

Wayne

noisephotographer · Mar 30, 2018

Wayne Larmon said:
DPReview has indicated that comparing cameras like the Pixel 2 to DSLRs using the DPReview test scene is problematic.

Wayne

Dpreview's comment was related to a user that doesn't take iPhone photos with Apple's app. The studio comparison is problematic when you don't consider the multi frame image processing. But it was considered. The dpreview comparison was shot with the Google app. This way you only get one Iso example in low light and in the same light conditions a Canon camera + f/3.5 lens would have needed Iso 3200 for the same brightness/exposure (Google's exif is related to a single exposure, that's at least my experience). Dpreview itself used the studio comparison in order to compare the Pixel 2 noise/details with the Rx100.

But of course every scenario can be different. A studio comparison only compares the studio scene. It doesn't compare blown out sky, moving objects, camera shake, etc..

Wayne Larmon · Mar 30, 2018

noisephotographer said:
But of course every scenario can be different. A studio comparison only compares the studio scene. It doesn't compare blown out sky, moving objects, camera shake, etc..

Yes. This is possibly what DPReview's Rishi Sanyal was referring to when he said "It's a difficult undertaking but something we're determined to do." Blown out skies can be emulated with lighting. But it is more difficult to replicate comparing moving objects and camera shake. This implies some kind of automated way of moving and shaking subjects (or cameras.)

FWIW, I was real impressed with my Pixel 2's IS when I took a few videos. And with stills for that matter. Very few blurred Pixel 2 images.

Wayne

PhotoFactor · Mar 30, 2018

Wayne Larmon said:
Galeta said:

Hi!

While using DPreview camera studio scene comparison tool I was barely able to notice the quality difference between Pixel 2 XL and Canon SL1/100D.

Is this me who do not know where and how spot difference properly or there is not that much of it?

Click to expand...

It isn't you. I've been comparing my Pixel 2 against my Canon M6 for about a month. Some results were posted in this thread.

I haven't changed my opinion much since that thread. Which is that OTC Pixel 2 JPEGS are often comparable in quality to M6 raw files that were tweaked with CC 2018 ACR (described in the above-linked thread.)

What that thread doesn't cover was the rest of the pictures I took on vacation. Because it was a vacation I didn't spend a lot of time taking matching Pixel 2/M6 images. But when I was going through my images after I got home I found that M6 images weren't appreciably better than Pixel 2 images. If both cameras captured the same scene and the Pixel 2 got a shot I wanted, I went with the Pixel 2 and didn't attempt processing the M6 raw.

Mainly the only time I used M6 images were when I needed more reach (55-200mm lens) than I could get with zooming-by-cropping Pixel 2 images. Or wanted the 10mm focal length I got with my 10-22mm lens. Or the two or three images that might be wall hangers and 24 megapixels were needed.

Otherwise clean 12 megapixel Pixel 2 images more than sufficed for vacation pictures, displayed on the highest resolution monitor I have (currently my 2160x440 Surface Pro 3.) Zooming into the images indicates that Pixel 2 images would hold up at higher resolutions. (4K is about 8 megapixels...)

PM me if you want a link to these images.

Wayne

Hi Wayne, glad to hear the phone continues to work well for you and impress!

Now you know why I pretty much am going to buy Google phones from now on. The camera, combined with the always-up-to-date Android version and security patches.

noisephotographer · Mar 30, 2018

Yes, but the OIS isn't really the main reason why you get sharp photos. It's mainly HDR+ which can replace OIS (it's even better than OIS in my opinion) because HDR+ takes multiple exposures with a fast shutter speed and picks the sharpest frames. That's my Google Nexus 5x experience and the Nexus 5x doesn't have OIS. The first Pixel doesn't have OIS, too. Only in low light it can get problematic when HDR+ chooses 1/10s. Then the OIS should help.

Wayne Larmon · Mar 30, 2018

noisephotographer said:
Yes, but the OIS isn't really the main reason why you get sharp photos. It's mainly HDR+ which can replace OIS (it's even better than OIS in my opinion) because HDR+ takes multiple exposures with a fast shutter speed and picks the sharpest frames. That's my Google Nexus 5x experience and the Nexus 5x doesn't have OIS. The first Pixel doesn't have OIS, too. Only in low light it can get problematic when HDR+ chooses 1/10s. Then the OIS should help.

The way that Google Research described the capture process is

Burst size In addition to determining exposure time, gain, and dynamic range compression, we must also decide how many frames to capture in a burst. The number we capture,N, is a tradeoff. In low light, or in very high dynamic range scenes where we will be boosting the shadows later, we want more frames to improve signal-to-noise ratio, but they take more time and memory to capture, buffer,and process. In bright scenes, capturing 1–2 images is usually sufficient, although more images are still generally beneficial in combating camera shake blur. In practice, we limit our bursts to 2–8 images, informing the decision using our model for raw image noise (see section 5 for more detail)

and

Reference frame selection To address blur induced by both hand and scene motion we choose the reference frame to be the sharpest frame in a subset of the burst, according to a simple metric based on gradients in the green channel of the raw input. This follows a general strategy known as lucky imaging [Joshi and Cohen 2010]. To minimize perceived shutter lag, we choose the reference frame from the first 3 frames in the burst.

From pages 4 and 5 of

Burst photography for high dynamic range and low-light imaging on mobile cameras

They then describe how the frames are broken down into thousands of tiles and each tile is compared to the reference image, mostly for alignment. "Alignment" is what keeps the merged image sharp. They move the tiles around so they match the reference image (if possible) before merging.

This paper was from 2016 and predates either Pixel. They only mention Nexus models. So it doesn't address OIS. In any event, the process works well.

You probably know all this because I got all my Google Research references from you, from earlier posts.

Wayne

Wayne Larmon · Mar 30, 2018

PhotoFactor said:
Hi Wayne, glad to hear the phone continues to work well for you and impress!

Now you know why I pretty much am going to buy Google phones from now on. The camera, combined with the always-up-to-date Android version and security patches.

Google has certainly raised the bar, but now all the smartphone companies are competing on camera quality. (Which is a good thing.) DPReview hasn't reviewed the Samsung S9 yet.

I don't know how high a priority Google has with camera performance in the Pixels (or whatever they will call them in the future). Google's main interest in Android is increasing market share. So if another company that makes Android phones pulls ahead, Google might not care. So long as more Android phones are sold.

Maybe. This is all pure speculation on my part. All I really know is that the more I use my Pixel 2, the more I am impressed with the image quality. As I've posted elsewhere, (resolution aside) usually images I get from my M6 raw files don't beat my Pixel 2 by much.

Wayne

noisephotographer · Mar 31, 2018

Wayne Larmon said:
noisephotographer said:

Yes, but the OIS isn't really the main reason why you get sharp photos. It's mainly HDR+ which can replace OIS (it's even better than OIS in my opinion) because HDR+ takes multiple exposures with a fast shutter speed and picks the sharpest frames. That's my Google Nexus 5x experience and the Nexus 5x doesn't have OIS. The first Pixel doesn't have OIS, too. Only in low light it can get problematic when HDR+ chooses 1/10s. Then the OIS should help.

Click to expand...

"Alignment" is what keeps the merged image sharp.

Yes, but nonetheless it requires a fast shutter speed for the best results.

Wayne Larmon said:
This paper was from 2016 and predates either Pixel. They only mention Nexus models. So it doesn't address OIS. In any event, the process works well.

Yes, the Pixel 2 has OIS additionally.

PGH423 · Apr 2, 2018

Galeta said:
Hi!

While using DPreview camera studio scene comparison tool I was barely able to notice the quality difference between Pixel 2 XL and Canon SL1/100D.

Is this me who do not know where and how spot difference properly or there is not that much of it?

Are we looking at the same thing? (I'm honestly not sure - I've never looked at these comparisons before.) To me, the Pixel 2 pictures look slightly sharper but MUCH noisier. Here is one screen shot:

Wayne Larmon · Apr 2, 2018

PGH423 said:
Galeta said:

Hi!

While using DPreview camera studio scene comparison tool I was barely able to notice the quality difference between Pixel 2 XL and Canon SL1/100D.

Is this me who do not know where and how spot difference properly or there is not that much of it?

Click to expand...

Are we looking at the same thing? (I'm honestly not sure - I've never looked at these comparisons before.)

When comparing smartphones that use heavy computational photography against conventional cameras that don't it is difficult to set up conditions that are comparable. Yesterday I did a few comparison shots of my Pixel 2 vs. my Canon M6 in natural light. One Pixel 2 shot however Google auto-set everything and three M6 shots at different apertures. Hover over the pics to see the EXIFs

Pixel 2

Look at them all at "original size". When the M6 is at low ISOs (large aperture, shallow DOF) the sensor looks good. When the ISO goes up (narrower aperture, deeper DOF), the image looks gnarly. (Any M6 blur is because of camera shake. These images also aren't for demoing sexy bokeh--just noise comparisons at different M6 ISOs.)

Because the Pixel 2 has a tiny sensor its DOF is deep. Google's computational photography effectively multiplies the sensor's size by nine so it is almost equivalent to a M/3 sensor noisewise. (See above example, at 100%.) Conventional cameras that do large sensor sizes brute force are larger and the cameras pay the price with intrinsic shallow DOF that can't be fixed in PP. Cameras that use computational photography (Pixel 2, iPhone X and most other newer flagship phones) can compute as much exquisite shallow DOF as you want.

And only on the portions of the image that you want. Dial in an Otus. So sayeth DPReview.

You pays your money and you takes your choice.

Wayne

Jostian · Apr 4, 2018

Wayne Larmon said:
PhotoFactor said:

Hi Wayne, glad to hear the phone continues to work well for you and impress!

Now you know why I pretty much am going to buy Google phones from now on. The camera, combined with the always-up-to-date Android version and security patches.

Click to expand...

Google has certainly raised the bar, but now all the smartphone companies are competing on camera quality. (Which is a good thing.) DPReview hasn't reviewed the Samsung S9 yet.

I don't know how high a priority Google has with camera performance in the Pixels (or whatever they will call them in the future). Google's main interest in Android is increasing market share. So if another company that makes Android phones pulls ahead, Google might not care. So long as more Android phones are sold.

Maybe. This is all pure speculation on my part. All I really know is that the more I use my Pixel 2, the more I am impressed with the image quality. As I've posted elsewhere, (resolution aside) usually images I get from my M6 raw files don't beat my Pixel 2 by much.

Wayne

The other cool thing is that the Google Camera ports that are out there allow for non Pixel phones to get the Google Camera IQ on their phones, its really great, I compared it on my S8 (the OEM camera app vs the GCam port (Google Camera port) and the difference (as good as the IQ was from the S8) was easy to see. Similarly with the LG G6 see below, the G6 is decent IQ wise (crop on left side), but the GCam port crop (right side) shows a much more natural (non over-processed) look.

So on my Essential Phone I use the Google Camera port app for all photos except for B&W shots, for those I use the OEM camera app for (has a dedicated monochrome sensor) and results are great.

--
Jostian

https://www.flickr.com/photos/jostian/

Search

Pixel 2 XL vs Canon SL1/100D

Galeta

New member

noisephotographer

Guest

Wayne Larmon

Forum Pro

noisephotographer

Guest

Wayne Larmon

Forum Pro

noisephotographer

Guest

Wayne Larmon

Forum Pro

PhotoFactor

Veteran Member

noisephotographer

Guest

Wayne Larmon

Forum Pro

Wayne Larmon

Forum Pro

noisephotographer

Guest

PGH423

Leading Member

Wayne Larmon

Forum Pro

Jostian

Veteran Member

Similar threads

Keyboard shortcuts