Canon 70D - Impressions, Discussion, etc.

[neonblingbling]

I'm lost; are you two experimenting with how the cameras will set exposure (70D needs to be set 'darker')?

I want to say from hokie's radio picture that the black from the radio boxes is throwing off exposure. My guess is if Dark takes the lens picture again with the TV as a backdrop, it might require the -1.3ev for the exposure to look 'right' since the black TV screen will throw off the exposure, even though lighting hasn't changed.

EDIT: Hokie, dumb question, but are you using the same metering mode in both experiences?

 

[hokiealumnus]

I'd say what you see is about right. Unfortunately, by the time I compress and upload them, the difference isn't as dramatic as what I'm seeing.

Your picture profile, efs 18-55@ 23mm, iso 1600, f3.5. Light source is a little 13w cfl spotlight with a diffuser in an otherwise dark room.

ev -.7
View attachment 137597
eV -1.3
View attachment 137598

Thanks DnB, that looks about right to me. The -1.3 eV seems like it might be more accurate, yes?

Also, I see a red ring on that lens; very nice.

I'm lost; are you two experimenting with how the cameras will set exposure (70D needs to be set 'darker')?

I want to say from hokie's radio picture that the black from the radio boxes is throwing off exposure. My guess is if Dark takes the lens picture again with the TV as a backdrop, it might require the -1.3ev for the exposure to look 'right' since the black TV screen will throw off the exposure, even though lighting hasn't changed.

EDIT: Hokie, dumb question, but are you using the same metering mode in both experiences?

Yes, evaluative metering. Someone at another forum (I've posted this here, OCN, XS and FM to get a variety of feedback) mentioned I should try different metering, which makes sense. Evaluative works great most of the time; this is a very specific situation and it seems to be worse with flash.

So yes, we're experimenting with how they set exposure, specifically in dark conditions. The 70D over-exposes when it's dark (artificial, indoor lighting).

 

[Dark_n_Beyond]

Thanks DnB, that looks about right to me. The -1.3 eV seems like it might be more accurate, yes?

Also, I see a red ring on that lens; very nice.

Yep... -1.3 eV is about as close as it gets. I guess it's not too surprising when you consider that our camera sensors are much more sensitive to light than our eyes.

As for the red ring... it's a great conversation piece... for a coffee cup. Don't feel bad, it fools everyone who hasn't seen it with the cover off or picked it up.

 

[hokiealumnus]

Haha...and I almost got one of those cups too, which makes that even more embarrassing.

 

[ASUS ROG]

I don't have any comment about the camera but you have a very smart looking kid, maybe when he's grow up, he will beat his dad in the overclocking game

 

[hokiealumnus]

Thanks! I'm quite sure he'll be there in no time. He's already started enjoying this hobby though, thrilled as he can be when I let him use the SLR. Truth be told, I prefer he use the point & shoot though, his little hands can barely hold this thing.

Another observation as far as the camera goes: Over the weekend I put the 18-135mm f/3.5-5.6 IS on and turned on live view video, then panned around the room. What a racket that thing made! Granted, it was quiet in our house, but even if I did put a shotgun mic on there, I can't imagine it would silence that sound.

The lens doesn't strike me as loud, but that's because I've always shot photos with it, where it focuses and stops there. For video with constant focusing adjustments, it's completely unusable.

Soooo, it looks like I'll be limited to 18-55mm when filming video for now (which isn't really that bad; on a crop sensor that's 88mm equivalent). I put my ear up to the STM lens when panning the camera around and could barely perceive a tiny little sound. It's perfect for video, even with the built-in mics. Thus, if you want video with more range, I'd recommend getting the kit with the 18-135mm STM lens.

 

[hokiealumnus]

One thing I've noticed messing around in the house (it's rainy here) is that this camera over-exposes in dark'ish indoor conditions, such as a room with couple of lamps, or a shop with a couple lights in the ceiling. I have to adjust exposure compensation down to between -1.3 to -2.0 eV to get good images. It takes great images when adjusted of course, but I'm surprised how far down it has to go. The T3 usually took ~-2/3eV in similar conditions. It's no problem - all cameras to some extent try to over-expose in dark conditions, but the amount of adjustment required was unexpected.

Ok then, I sheepishly must say that this was all my fault. There's this new setting called "Highlight Tone Priority", which is supposed to bring detail back to blown-out highlights, which would normally be a featureless white blob in your photo.

When HTP is turned on, that automatically turns off Auto Lighting Optimizer. I'm used to shooting with ALO on and have gotten used to its behavior on the T3. Well, with it disabled, plus the camera trying to recover highlights in a dark environment (indoors), it blew out the majority of the image.

Once I turned HTP off and with ALO back on, the camera started acting like I expected it to again, slightly over-exposing as expected with ~-0.3eV required in low light conditions.

The moral of the story - don't turn on features you aren't familiar with and forget they're on. Wait, no, that's not it. The moral of the story is that in low light situations, indoors with low'ish lighting, HTP will indeed give highlights the priority and blow out the rest of the photo.

 

[Dark_n_Beyond]

Now I'll have to check my settings. I'm pretty sure I never turn ALO on, and I don't think I've ever use HTP either. On the other hand, I play with settings so much, no telling what's on or off at the moment. Thanks for setting it straight though.

 

[hokiealumnus]

ISO Tests!

ISO Noise Testing

These ISO tests were all taken with my standard image profile: Auto picture style, +6 sharpening, +1 saturation. The bookshelf is in our bonus room / office. The only light source was a 100W equivalent, daylight CFL in a fan, in the middle of the ceiling. The light source is approximately five feet away and about 1.5-2 feet above the shelf you're seeing.

To keep variation between shots minimal, they were taken with spot metering (center of the frame) and a single AF point (which is in the middle of the metering spot). The lens was the kit lens (18-55mm STM) at f/4.5 & 35mm

I'll split this into two posts to prevent bumping up against any potential per-post image limits. The first post is ISO 100 through ISO 800, the second is ISO 1600 through ISO 25600.

ISO 100

ISO 200

ISO 400

ISO 800

Everything looks fine through ISO 800 when considering the full images.

 

[hokiealumnus]

ISO 1600

ISO 3200

ISO 6400

ISO 12800

ISO 25600 (aka zomg so much noise!)

ISOs 1600 through 6400 seem quite usable. 1600 has some noise, but it's not very intrusive. At 3200 and even 6400 you can start to see a bit more but it's not bad.

When you start going to ultra high ISOs, 12800 starts getting pretty noisy. I'd probably not use that unless I absolutely needed the shutter speed and there was no other choice. ISO 25600 is really just a gimmick or for photo emergencies when it's pitch black, you have no tripod and you absolutely must take a photo right then. Aside from this test, I don't even have that ISO enabled for selection.

Now, I have not cropped any of these images (still haven't decided which part of the image to crop) nor really had time to look at them closely; so consider these just plain raw data for you to draw your own conclusions. Taken as full images, all the way up through ISO 6400 looks great to me. ISO 12800 is even ok, but not preferable. ISO 25600...just no.

If anyone has any suggestions about which part(s) of the images to crop that would be the most challenging for comparison (two or three is fine) at 100%, I'm all ears.

 

[Dark_n_Beyond]

I would say probably the 3 wood uprights in the top right corner. Typically you'll see loss of detail in the light neutral tones first.

edit: based on my first observations, 1600 looks pretty good, 3200 acceptable, 6400 is probably higher than I'd personally go.

 

[hokiealumnus]

Thanks, I'll look at that.

Heh, my 6400 take is colored heavily by the T3, which looked at 3200 about what the 70D looks like at 12800.

 

[GTXJackBauer]

Not to ruin the party but with the lower ISO the sharper it looks? (At least to me lol) The lower ISO is best used for light/sun and faster shutter speed. Higher ISO used for Night with a slower shutter speed?

Why would you go higher ISO on a light image that looks really good at 200? Is there a point to go to the highest ISO for example on yours say 1600 than the 200? Higher ISO pushes more colors than lower?

Sorry. I have to ask or I'll just always wonder. Plus I get to run back to these and see you guy's answers to keep me in line.

Figure these are like my "notes". Actually I should be taking these bits of info and writing them down.

 

[hokiealumnus]

Lower ISO will always be cleaner. ISO is the sensor's sensivity setting. The higher the ISO, the more sensitive it is to light. Unfortunately that sensitivity comse at a price - noise.

You always want to shoot at as low an ISO as you can for what you're shooting. The problem with that (and this is where you're getting confused) is that as light goes down, the shutter has to stay open longer if the sensor is at its lowest sensitivity (ISO 100 typically).

If you are shooting hand held, when it gets darker, that shutter speed is a problem. As it is if you are shooting things like sports with lower lighting. You run the risk of a blurry exposure (camera shake) or blurry subjects (movement while the shutter is open).

To compensate, you increase the sensitivity of the sensor (higher ISO) in order for it to allow for less light coming in (faster shutter).

As you can see above, higher ISO means noise, hence the advice to usethe lowest ISO you can. Real world example: That shot aabove at ISO 100 never could have happened without a tripod. Nobody's hands could hold a camera perfectly still that long.

 

[GTXJackBauer]

Lower ISO will always be cleaner. ISO is the sensor's sensivity setting. The higher the ISO, the more sensitive it is to light. Unfortunately that sensitivity comse at a price - noise.

You always want to shoot at as low an ISO as you can for what you're shooting. The problem with that (and this is where you're getting confused) is that as light goes down, the shutter has to stay open longer if the sensor is at its lowest sensitivity (ISO 100 typically).

If you are shooting hand held, when it gets darker, that shutter speed is a problem. As it is if you are shooting things like sports with lower lighting. You run the risk of a blurry exposure (camera shake) or blurry subjects (movement while the shutter is open).

To compensate, you increase the sensitivity of the sensor (higher ISO) in order for it to allow for less light coming in (faster shutter).

As you can see above, higher ISO means noise, hence the advice to usethe lowest ISO you can. Real world example: That shot aabove at ISO 100 never could have happened without a tripod. Nobody's hands could hold a camera perfectly still that long.

Oh so the trade off is the higher ISO is best for fast shots but will become grainy where the lower ISO shots are longer shutters but prone to blurry images not on a tripod. I've noticed that with my camera and didn't get it why it took long on some and fast on others. Best images came out when I had it on a tripod.

So is it safe to say for fast shutters, say a thunderstorm outside and want to catch some clouds or lightening, I should put the ISO higher since I won't be on a tripod but mobile.

For some flower shots I took in my yard during summer of last year, the shutter was pretty long and stood still as much as I could. The shots came out amazing.

The longer the shutter the better the quality? It gives the sensor time to absorb?

Thanks for the clarification hokie.

 

[neonblingbling]

If anyone has any suggestions about which part(s) of the images to crop that would be the most challenging for comparison (two or three is fine) at 100%, I'm all ears.

I would personally try somewhere that has lots of fine detail that should be in focus. Perhaps the gray Manstein book (it looks like fabric)? I'm curious to see how aggressive the noise reduction is.

 

[Dark_n_Beyond]

I would personally try somewhere that has lots of fine detail that should be in focus. Perhaps the gray Manstein book (it looks like fabric)? I'm curious to see how aggressive the noise reduction is.

I looked at that same book quite a bit too, the closest thing to fabric in the picture. I can agree with that.

 

[hokiealumnus]

The longer the shutter the better the quality? It gives the sensor time to absorb?

You start off great through the next to last sentence (quoted), but there's where you slip up. Shutter speed itself doesn't directly relate to quality. I would say the lower the ISO, the better the quality. It is your lighting, in conjunction with your ISO, that will determine whether your shutter speed is long or short.

Even at ISO 100, if you have a long shutter speed in broad daylight, your exposure will be blown out (over-exposed). Likewise if you're in moderate light, but have set a high ISO (more sensitive to light) and your shutter is open too long, it will also be blown out.

I'll give the ISO & shutter speed for all of the shots above, hopefully you can start to see the relation.

ISO - Shutter Speed (in seconds)
100 - 1.3
200 - 0.6
400 - 1/3
800 - 1/6
1600 - 1/13
3200 - 1/25
6400 - 1/50
12800 - 1/100
25600 - 1/200

Changing your ISO a "full stop" (above are full stops, you can set 1/3 or 1/2 stops too, depending on the camera; I still haven't gotten used to seeing numbers like ISO 2000 on the new camera yet) amounts to you telling the camera's sensor to double its sensitivity to light, hence the doubling of number equating to a full stop (see, it's logical!).

As your sensor doubles in sensitivity, your shutter likewise cuts the time it needs to be open by a proportionate amount. All other things being equal, as they were in these shots, doubling the ISO (sensitivity) equates to cutting shutter speed in half (amount of light let in).

Hope that makes sense.

DnB and neon, thanks for the suggestion, I'll get some close crops later on. I do have one closer crop I'll post sometime soon. It's not that close-in though, your guidance points me to good places to get closer.

 

[hokiealumnus]

ISO Tests Cropped

These are the first crops of the ISO tests above. I will be cropping tighter in at least one more place, if not more than that, as time allows. These are not quite as viewed at 100%, but quite close; they were resized by ~120px horizontally. I'll also split these into two posts.

ISO 100

ISO 200

ISO 400

ISO 800

One spot to watch in this series is the book on the far left. It has a fabric binding and will start to lose detail pretty quickly when the camera fights against high ISO noise.

As far as this series of four goes, things look pretty solid through ISO 800, but you can definitely see a loss of detail starting at ISO 400 and becoming more noticeable at 800.

 

[hokiealumnus]

ISO 1600

ISO 3200

ISO 6400

ISO 12800

ISO 25600

ISO 1600 still looks ok for detail, but you're starting to get more visible noise. 3200 has more visible noise, but it starts to add a degree of softness that wasn't there at 1600 and below. Notice the "Richmond" text in the binding second from the left.

At 6400 you're getting a good big of noise through in addition to a very noticeable drop in sharpness from 3200. 12800 is just plain noisy and the camera's algorithms are struggling to combat the noise, even as the image becomes softer. It's doing a good job, but again this is where it starts to become unusable, depending on the shooting conditions.

As expected and already noted, 25600 is just there to say the camera has that setting. It's straight up ugly, with little discernible sharpness and lots of noise.