Impress me with high ISO

[dod]

Okay, there are one or two in this thread that seem to like arguing just for the sake of it.

Before you post anymore, please take a second to think about what you're saying and how it could be interpreted. If there's a chance it's going to wind people up, don't post it, find an alternative way of saying the same thing.

 

[Pookeyhead]

I don't really see the point in showing off camera shots - we don't show these to others...?

But... if anyone is interested in shooting in low light, they will want to see how well the camera alone can cope with it, not how well you've dealt with it in post processing. Yes.. you deal with noise in PP regardless of how good your camera is, but the better the camera is raw, the better you can make it in PP. It's still perfectly valid in a thread like this to post up unedited RAW.. and at full res so we can pixel peep. This is talk equipment... not the post processing forum, or the image crit thread forum.

If I was looking for a new camera for this purpose, I'd come in here expecting lots of practical, real word examples of how various cameras perform, not various examples of how good people are at reducing noise in PP.

I hope that made sense... not had any coffee yet

 

[Gary Coyle]

But... if anyone is interested in shooting in low light, they will want to see how well the camera alone can cope with it, not how well you've dealt with it in post processing. Yes.. you deal with noise in PP regardless of how good your camera is, but the better the camera is raw, the better you can make it in PP. It's still perfectly valid in a thread like this to post up unedited RAW.. and at full res so we can pixel peep. This is talk equipment... not the post processing forum, or the image crit thread forum.

If I was looking for a new camera for this purpose, I'd come in here expecting lots of practical, real word examples of how various cameras perform, not various examples of how good people are at reducing noise in PP.

I hope that made sense... not had any coffee yet

Going back a tad to where you mentioned an 800px wide sample not really giving us enough to go on, i dont mind unedited samples that size as on one of my screens that size approximates roughly to what a 7 inch print would look like.

I dont shoot Raw, never needed to, never felt the inclination to do so either, But would a RAW image converted to JPeg with no PP give me an image with less noise than a one shot in Jpeg in the first place.

 

[Pookeyhead]

Going back a tad to where you mentioned an 800px wide sample not really giving us enough to go on, i dont mind unedited samples that size as on one of my screens that size approximates roughly to what a 7 inch print would look like.

I'd want to load it up in PS and look at it 100% and make my own judgements on camera outputs before I bought one. Would you buy a camera if all you had to evaluate it was a 7 inch print? I think most people would want to critically examine it.

I dont shoot Raw, never needed to, never felt the inclination to do so either, But would a RAW image converted to JPeg with no PP give me an image with less noise than a one shot in Jpeg in the first place.

That's up to you, but JPEGs may have had noise reduction applied in camera, whereas RAW files have no processing applied, so again, if I wanted to critically examine a shot for noise prior to purchasing it, I'd be looking at an unedited RAW file that I can be sure no work had been done to. That will all be in the EXIF data too, so I'd be suspicious of images that have the metadata removed, or absent.

I'm in a fortunate position of being able to get my hands on, and use most new cameras before buying them so I can make my own evaluations... most aren't, and a thread full of unedited, full resolution images would be something I'd be glad I'd found if I was trying to decide between cameras.

 

[Gary Coyle]

I'd want to load it up in PS and look at it 100% and make my own judgements on camera outputs before I bought one. Would you buy a camera if all you had to evaluate it was a 7 inch print? I think most people would want to critically examine it.



That's up to you, but JPEGs may have had noise reduction applied in camera, whereas RAW files have no processing applied, so again, if I wanted to critically examine a shot for noise prior to purchasing it, I'd be looking at an unedited RAW file that I can be sure no work had been done to. That will all be in the EXIF data too, so I'd be suspicious of images that have the metadata removed, or absent.

I'm in a fortunate position of being able to get my hands on, and use most new cameras before buying them so I can make my own evaluations... most aren't, and a thread full of unedited, full resolution images would be something I'd be glad I'd found if I was trying to decide between cameras.

Trouble is the "normal" guy on the street doesnt get a chance to evaluate full size unedited images before buying which he himself has shot, imagine the look on the face of the Jessops sales guy if you asked him if he could turn the lights off in the store while you tried the brand spanking new D4 out for a while to see how it handles high ISO, he would melt on the spot, only real solution is to buy online, test it at home then send it back under DSR if not happy, im not advocating this by the way although it is something i have done with several pieces of Electronic/Televisual/Video equipment. (not for noise) If a manufacturer says their product does something but i think it doesnt deliver then it goes back.

But yes youre right, if i personally were spending 4k on the next big thing i would test it out myself, i know i harp on about the online reviewers not testing real world situations and thats why i would want the camera in my sticky mits to test it first, i shoot in dark dungeon like stadiums so need to test in those places and not in a nicely lit studio environment.

I will deffo give this RAW business a try out sometime (but not pitchside), i suppose it really must be the very best way to take photos.

 

[Deleted member 49549]

ISO3200 - same level as noise as 16000 on your 5D2.

Its not just about noise - dynamic range is something that is impacted by high iso, see the amount of blow highlights in the above pics, or were they overexposed to counteract noise?

 

[Phil Young]

But... if anyone is interested in shooting in low light, they will want to see how well the camera alone can cope with it, not how well you've dealt with it in post processing. Yes.. you deal with noise in PP regardless of how good your camera is, but the better the camera is raw, the better you can make it in PP. It's still perfectly valid in a thread like this to post up unedited RAW.. and at full res so we can pixel peep. This is talk equipment... not the post processing forum, or the image crit thread forum.

If I was looking for a new camera for this purpose, I'd come in here expecting lots of practical, real word examples of how various cameras perform, not various examples of how good people are at reducing noise in PP.

I hope that made sense... not had any coffee yet

Fair point that.

Still, for the benefit of the OP, I still think it would be a nice inspiration to see what he could get should he venture beyond ISO1000 and apply some PP if required.

I'm not convinced 100% crops to show noise are the way to go in this particular case as the op obviously is trying to stay away from noise like its the devil...showing the devil's face won't help things...

I could be wrong...but I would hazard a guess that ISO3200 is the same on both 5D models (or at least very close) and so OP can gain at least one stop of light just by pushing things a bit more now.

 

[Phil Young]

Its not just about noise - dynamic range is something that is impacted by high iso, see the amount of blow highlights in the above pics, or were they overexposed to counteract noise?

Not disputing that at all...OP seems mainly concerned about noise though.

The blown highlights are just light trails so you would expect a level of over exposure anyway but in this case I increased the highlight levels to give effect.

 

[wyx087]

the real issue here is that OP is cropping way too much.

I've test shot and printed 7x5 from 3200 to 25600. 3200 and 6400 without noise reduction it's impossible to tell apart. at extended H1 12800 with slight noise reduction it's as good as 3200, but dynamic range suffers slightly. at 25600 it's passable, but dynamic range makes the shot not really worth the effort.

so with my 5D2 I trust auto-ISO at 3200 max, and never think about it. when lighting is low, I won't hesitate to go to 6400. when only for web (most of my shots) I won't hesitate to go to 12800.

OP just need to get a longer lens. the camera body is perfectly fine. don't pixel peep at 1:1, don't crop too much and don't be afraid to use some noise reduction.

 

[artyman]

All this talk about high ISO's so I tried a quick shot with my 7D, 12800, looks acceptable to me.

 

[Deleted member 49549]

Not disputing that at all...OP seems mainly concerned about noise though.

The blown highlights are just light trails so you would expect a level of over exposure anyway but in this case I increased the highlight levels to give effect.

On my monitor there are a lot of areas overexposed / blown! What is the exif for the photo and what is the PP? ie did you overexpose the initial shot and then pull back in pp?

 

[Pookeyhead]

Trouble is the "normal" guy on the street doesn't get a chance to evaluate full size unedited images before buying which he himself has shot,

Exactly... which is why I think people SHOULD post up unedited full res images to evaluate.

 

[Gary Coyle]

Exactly... which is why I think people SHOULD post up unedited full res images to evaluate.

You (not you personally) can always use pixel peepers website to choose any camera and lens combo and choose to link to full size inages hosted on FlickR, its not perfect but gives a good idea, you just need to hope theyre unedited and EXIF intact, better than nothing though.

 

[Ralphmyster]

Cheers guys but remember this is sports with moving subjects and moving water I am well aware how good the 5dmkii is BUT in my opnion I wanted better. And to see if there is a camera that betters it. I would go for a 400mm lens if i was shoot lots and earning lots LOL I am doing a shoot with Rebecca Adlington next week so will take a few with high iso and see if I like them I do like to have in my mind this image quality is the best it can be,

 

[Phil Young]

On my monitor there are a lot of areas overexposed / blown! What is the exif for the photo and what is the PP? ie did you overexpose the initial shot and then pull back in pp?

Here is the original (my monitor is caliberated and looks well exposed - bar the blown highlights). As you can see, it was less "blown" but I wanted to pull back even more. Anyway - bear in mind, it's not an image to critique we are all discussing ISO noise here...DR and white clipping has never been an issue raised

When shooting high ISO I expose more to the shadows. Sometimes this means the whites are clipped slightly but it enables me to get the best image in PP.

Exif:
f1.5
1/3s
ISO800

 

[Pookeyhead]

You (not you personally) can always use pixel peepers website to choose any camera and lens combo and choose to link to full size inages hosted on FlickR, its not perfect but gives a good idea, you just need to hope theyre unedited and EXIF intact, better than nothing though.

All true.. it's wondering if they've been messed with though. I suppose you'd have that problem on here though, as inevitably some will see start posting up edited images to try and "win". :bang:

 

[Nod]

Surely all that really matters is how the final image looks when it's in its final state in the final medium? If it's a crop from the central 600 x 400 pixels off the sensor, it may look OK at screen size but won't take much zooming before pixelation starts and any noise present becomes obtrusive. At that sort of zooming, most NR software will have reduced detail too. Not ideal! Print the same image at 6x4" and it may well be fine (unless peered at through a magnifier!) even at a mere 100dpi print resolution.

At the moment, I'm using my Fuji XF1 quite a lot, with its tiny (compared to my D700's) sensor - you know what, SOOC, with no sniff of PSE etc, ISO 1600 shots at A3+ are great! On screen, pixel peeping, of course they show some noise (examples are posted in a thread somewhere) but I dislike using screens as a viewing medium - there's far too much variation between monitors so colours, brightness etc are usually not what the photographer intends people to see.

Compared to film grain (which I love in pushed film B&W prints almost as much as I hate it in old, fast colour emulsions), most modern cameras' handling of high ISOs is verging on the incredible! The only things to remember are not to crop in too far and to be careful with NR and sharpening. Canon, Nikon, Pentax, Sony etc? Who gives a Castlemaine? Sony make most of the sensors anyway and ALL the brands make damn fine cameras. If you're finding then you need to crop in too far and noise is a problem, it's time to dig deep and get that longer fast lens you need (or, if the images are for your own consumption, just want!). Not an easy afford but the only real way to get that extra quality you're after.

 

[Pookeyhead]

Surely all that really matters is how the final image looks when it's in its final state in the final medium?

It's important yes, but if you're trying to decide between 2 cameras, you need to see what the CAMERA is capable of, not the photographer's editing skills, or the effectiveness of a NR process. Yes, you can make any camera's high ISO output much better with NR and of course you should, but start with a better camera and the results using the same process will be better still.

 

[Nod]

Should? Why? I loathe doing PP and avoid it whenever possible. As I've said, SOOC images from all the cameras I use are great at A3+ up to at least ISO 1600. If a camera/lens system delivers what is necessary for an image's final use, what's the point in searching for something to spend (often the bank's) money on?

 

[ernesto]

If a camera/lens system delivers what is necessary for an image's final use, what's the point in searching for something to spend (often the bank's) money on?

You do realise you are in the Talk equipment section don't you

I long ago realised that an entry level DSLR camera with a budget prime on the front gives me images which are as good as they need to be. That doesn't stop me thinking about gear though...

 

[Nod]

Me neither! Luckily, I can afford to indulge myself when I want to but equally luckily, I am also aware that I'm happy with what I've got so have given up searching for the next thing to spend my money on. Too many people feel the need to upgrade their kit when the extra expense may well cause them financial hardship, so if the kit they have does what they need, why get into debt (or risk doing so) especially just before christmas!

By all means, if the OP can afford to go out and get a D5IV and a long, fast prime (and can justify it to themself and any SO there may be), then why not but they may well not be able to afford to do that, so shouldn't (and looking for excuses to do so doesn't help).

 

[Phil Young]

It's important yes, but if you're trying to decide between 2 cameras, you need to see what the CAMERA is capable of, not the photographer's editing skills, or the effectiveness of a NR process. Yes, you can make any camera's high ISO output much better with NR and of course you should, but start with a better camera and the results using the same process will be better still.

The op doesn't need a new camera. he needs to shoot higher ISO...

is there anyone here that thinks the op will see drastic changes in noise between a 5D2 & 3?? I don't have the cameras so can only hypothesize...

 

[Pookeyhead]

The op doesn't need a new camera. he needs to shoot higher ISO...

is there anyone here that thinks the op will see drastic changes in noise between a 5D2 & 3?? I don't have the cameras so can only hypothesize...

I think you're right.

Anyway... as part of another thread, I hosted some straight RAW conversions to JPEGs from a range of cameras. All straight, no processing.


Make your own mind up.

All at ISO12800

Nikon D3s

Nikon D4

Nikon D800

Canon 5D MkII

Canon ID MkIV


All modern dSLRs are pretty much of a muchness these days.

 

[Neilly]

Hi,

I've already posted this picture once tonight so one more time won't hurt I guess. 1DX + 24-105mm L lens was used - the tunnel is very poorly lit and I was at approx the mid point. Anyone who's been through the recently opened tunnel on the Tissington Trail at Millars Dale/Monsal Edge will know how dim it is in there.

Oh, I forgot the ISO........................51200! And also - there's been no PP on the picture, just resized for the forum.

TTFN,
Neil